Bystander and recipient roles reweight noise, privacy, and safety in drone deliveries

Bystander and recipient roles reweight noise, privacy, and safety in drone deliveries | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Bystander and recipient roles reweight noise, privacy, and safety in drone deliveries Maria Kamargianni, Yuerong Zhang, Fangqing Song, Yanghui Cao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9412457/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Drone delivery promises faster, cleaner logistics but raises concerns about noise, privacy, and safety that differ depending on how people encounter the technology — whether as recipients who benefit directly or as bystanders who bear externalities. Current research rarely distinguishes between these perspectives, overlooking how preferences may depend on the context of that encounter. Using a discrete-choice experiment in Poland (N = 1,001) in which the same individuals evaluated matched scenarios from both perspectives, we estimate a pooled multinomial logit model with role-specific interactions. Results show role-dependent trade-offs. Bystanders penalise noisier operations more than recipients in routine settings, but this gap vanishes in noise-sensitive locations. Meanwhile, camera preferences are strongly context-dependent: bystanders favour cameras at public delivery points but resist them near private ones, while recipients show weak camera sensitivity overall. Both roles value operator licensing and drone registration, with bystanders placing modestly greater emphasis on compliance. These findings offer guidance for tailoring drone service design and communication to different publics. Physical sciences/Engineering Biological sciences/Psychology Social science/Psychology Figures Figure 1 Figure 2 Figure 3 Introduction Drone delivery is increasingly proposed as a lower‑emission option for the last mile, particularly for small, time‑sensitive payloads. Life‑cycle analysis shows that, when carefully deployed (e.g., light packages, low‑carbon electricity), drone delivery can reduce energy use and greenhouse‑gas emissions relative to diesel trucks [ 1 , 2 ]. Especially in favourable scenarios for 0.5 kg packages, the corrected estimate can reduce up to ~ 54% GHGs than truck delivery [ 1 ]. Beyond emissions, drone delivery may improve service accessibility for populations that face barriers to conventional logistics. For instance, people with limited mobility or those living in areas poorly served by road-based transport [ 3 ]. However, operational benefits alone do not guarantee viable services. Drone delivery may only scale if operations are accepted in the places where flights occur, including acceptance by people who do not receive the drone deliveries [ 4 – 6 ]. Existing research on drone delivery acceptance has largely treated the public as a single group. Most studies ask respondents to evaluate scenarios exclusively as prospective users of the service (see, for example, [ 7 , 8 ]). Large official surveys also report acceptance at the level of citizens as a whole, focusing on noise, privacy, and safety without separating groups by encounter role [ 9 ]. Importantly, these roles are not fixed segments of the population: the same individual may be a recipient in one encounter and a bystander in the next. Our within-person design leverages this by asking each respondent to evaluate scenarios under both perspectives, so that role differences are identified within individuals rather than between groups. To our knowledge, no study has asked the same individuals to evaluate matched drone-delivery scenarios from both the recipient and bystander perspective and then quantified how this role shift changes attribute valuations. Within-person designs are important here because they isolate role effects from stable individual differences (e.g., risk aversion, technology affinity) that would confound between-group comparisons [ 10 – 12 ]. The idea that the same technology is judged differently by beneficiaries and cost-bearers has been explored in other infrastructure contexts. For example, in the transport domain, Wüstenhagen et al. [ 6 ]distinguish community acceptance from market acceptance precisely because affected non-users may weight impacts differently from customers. This gap limits the practical value of existing evidence: without knowing whether bystanders and recipients weight the same attributes differently, policymakers and operators lack the basis for role-sensitive service design. Deployment in everyday spaces confronts three salient externalities: noise, privacy intrusion, and safety risk [ 13 – 16 ]. For noise, laboratory and field evidence shows that drone sounds are perceived as substantially more annoying than comparable ground-vehicle or conventional aircraft noise at equivalent exposure levels, creating a psychoacoustic burden distinct from familiar transport sources [ 9 , 17 – 19 ]. Public health guidance stresses context-specific management of noise exposure, recognising it as a material risk factor for wellbeing [ 19 , 20 ]. For privacy, concerns centre on the perceived surveillance potential of onboard cameras, which citizens judge differently depending on whether a drone operates near private property or in a public space [ 21 – 23 ]. From the broader citizen perspective, privacy and security are consistently identified as central barriers to acceptance [ 22 ]. For safety, concerns relate to the credibility of operational oversight: whether operators are identifiable, drones are traceable, and regulatory compliance is enforceable [ 24 , 25 ]. Risk research links such governance signals to higher trust and lower perceived risk, especially for novel technologies and for observers who lack direct control [ 24 , 26 ]. In the drone context, studies have advocated registration and remote identification to enable accountability and enhance public trust [ 25 , 27 ]. These three dimensions map directly onto policy levers — noise routing, camera and data-use regulation, and licensing regimes; making it essential to understand how different people value each one [ 6 , 28 – 30 ]. We therefore treat noise, privacy, and institutional safeguards as a minimal set of attributes that capture how citizens experience drone operations in daily life, and examine how encounter role shapes their relative influence. Beyond the role distinction, the operational context of a drone flight is likely to moderate the perception of externalities. For noise, public-health guidance stresses context-specific exposure management, implying that sensitivity should be heightened near homes and hospitals; places where quiet is both expected and health-relevant [ 19 , 20 , 24 ]. Psychoacoustic evidence reinforces this idea that drone sounds are judged more annoying than comparable transport noise at equal exposure [ 14 , 17 , 18 ], and this penalty should be most salient where ambient noise is lowest and expectations of quiet are strongest *e.g., hospitals). For privacy, the evaluation of an onboard camera is likely to depend on whether the delivery occurs at a public location (e.g., a park or commercial area) or a private one (e.g., a residential doorstep), since privacy norms differ sharply across these settings [ 21 , 23 , 31 ]. We therefore model interactions between role and these place-level variables. In addition, we examine urban–rural differences as a robustness check, based on the self-reported living area from respondents. Urban and rural residents differ in baseline ambient noise, housing density, and proximity to potential flight paths, all of which may moderate externality perceptions [ 32 ]. Prior work has examined drone acceptance across geographical contexts but has predominantly focused on urban populations [ 33 , 34 ], whereas comparative urban–rural evidence remains scarce despite the considerable potential of drone delivery in underserved rural areas. We test whether the role-dependent patterns observed in the pooled sample replicate across urban and rural subsamples, strengthening the generalisability of the role-based account. To test these expectations, we designed a within-person discrete-choice experiment fielded in Poland (n = 1,001). This design follows established practice in stated‑choice experiments and links to standard logit foundations used for behavioural choice modelling [ 35 , 36 ]. Each respondent evaluated two blocks of hypothetical drone-delivery scenarios (i.e. one from the bystander perspective and one from the recipient perspective), enabling direct comparison of how the same individual values the same attributes under each role. We estimate pooled panel multinomial logit models in which a role indicator is interacted with each attribute, so that a single model yields separate coefficient estimates for bystanders and recipients. In plain terms, the interaction structure lets us ask: does switching from recipient to bystander change how much a person cares about noise, cameras, licensing, or registration? The experiment included additional scenario attributes that differed across blocks: recipients saw delivery cost and time, while bystanders saw flight topography. These role-specific attributes enter the model through role-gated terms that restrict their contribution to the relevant subset of tasks, so that all attributes shown to respondents are included in the primary specification. A mixed logit robustness check and urban–rural subgroup analyses are reported in the Supplementary Information. We use the R package Apollo for estimation and post‑estimation checks [ 37 ]. Using this design, we find that encounter role systematically reshapes how people evaluate drone-delivery services. Bystanders penalise noisy operations more strongly than recipients in routine locations (e.g., commercial areas), but this role gap vanishes in noise-sensitive settings such as homes and hospitals, where both roles converge on similar noise sensitivity. Camera preferences are strongly context-dependent: bystanders favour camera-equipped drones at public delivery points but resist them near private ones, whereas recipients show weak camera sensitivity across both contexts. Both roles value operator licensing and drone registration, with bystanders placing modestly greater emphasis on compliance. These role-dependent patterns are broadly consistent across urban and rural subsamples. Results We collected data in the Upper Silesian Region (GZM) in Poland during February and March 2022, yielding 1,001 valid responses with 13,013 evaluated choice tasks. Each respondent completed two blocks of discrete‑choice questions: one block consisting of six scenarios framed as a bystander exposed to a drone delivery, and one block consisting of seven scenarios framed as a recipient of a drone delivery. The survey took place in February–March 2022. A survey company was responsible for recruiting the participants in our survey based on specific sample requirements (age, gender, and type of living area based on the census data of the GZM Metropolitan area) to secure representativeness of our sample. Sample characteristics and patterns of daily postal practice are reported in Supplementary Information S1 (Supplementary Table S1 and S2). We estimated pooled panel multinomial logit models with three alternatives (1: drone service A, 2: drone service B, 3: Neither), pooling the two role blocks into a single panel. The attributes and their descriptions shown to respondents are shown in Table 1 . Attributes were coded as described in Supplementary Information S2 (please see Supplementary Table S3): quietness/noisiness is treated as a five‑icon ordered scale (per‑step increase in noisiness), on‑board camera is coded dummy coded as "Camera vs No camera", and operator licensing and drone registration are coded as binary safeguards using "Licensed vs Not licensed" and "Registered vs Not registered" indicators. Parcel content was recoded into medicine, food, and other (baseline). Context variables capture noise‑sensitive locations (homes/hospitals vs other areas) and private versus public delivery points. Recipient-only attributes (delivery cost and delivery time) and a bystander-only attribute (flight topography) enter through role-gated terms that restrict each to the relevant task block. Table 1 List of attributes and levels Parcel type Attribute levels (bystander) Attribute levels (recipient) Medicines, food, clothes, mail, illegal items, surveillance, organic material Medicine, food, clothes, mail, surveillance, illegal items Operator licensing With license, without license With license, without license Drone registration Registered, not registered Registered, not registered Drone camera With camera, without camera With camera, without camera Drone noise 5 levels (0–4) 5 levels (0–4) Delivery building Hospital, home, business/ office area, shopping area Hospital, home, business/ office area, shopping area Delivery point Window, doorstep, communal space (e.g., garden), private terrace, open public space Window, doorstep, communal space (e.g., garden), private terrace, open public space Delivery cost N/A Free, 10 zł, 20 zł, 30 zł, 40 zł, 50 zł Delivery time N/A 20mins, 30mins, 40mins, 50mins, 60mins, 120mins, 6h, 12h, 24h, 48h Topography Over city, over village, rural area N/A The primary model includes all attributes shown to respondents. Attributes common to both role blocks (i.e. noise, on-board camera, operator licensing, drone registration, and parcel content) enter utility with a recipient-baseline coefficient and an additive bystander-shift parameter, so that role differences are identified within individuals. Attributes that appeared in only one block (i.e. delivery cost and time for recipients, flight topography for bystanders) are incorporated through role-gated terms, so they are set to non-available for outside the relevant role. This specification preserves direct within-person comparability for the shared attributes [ 38 ]. The primary model specification includes role‑specific interactions: a binary indicator (bystander = 1, recipient = 0) shifts the slopes of all shared attributes and their context interactions so that differences between viewpoints are identified within individuals. In plain terms, this tests whether the same person values noise, cameras, or safeguards differently when imagining themselves as a bystander versus a recipient. A likelihood‑ratio test strongly favours this interaction model over a pooled main‑effects model without role interactions (ΔLL = 58.43; LR = 116.86 on 9 df, p < 0.001). Information criteria also favour the interaction model (AIC = 25,843.9 vs 25,942.8; BIC = 26,008.3 vs 26,039.9), and pseudo‑R² against equal shares rises from 0.094 to 0.098. Full coefficient tables and fit indices are provided in Supplementary information S3 (Supplementary Table S4). Because the recipient and bystander blocks may differ in random utility scale, we have also estimated an alternative pooled specification that allows a role-specific scale factor for bystander tasks, normalising the recipient scale to one [ 39 , 40 ]. The estimated bystander scale parameter was not significantly different from unity, and the log-likelihood improvement was negligible, indicating that a common-scale specification is adequate (see Supplementary Information S3 for details). The scale factor is not separately identified once role-specific slope shifts are included, so we report the common-scale specification as our main model. The following sub-sections report role-gap odds ratios for noise, cameras, and institutional safeguards across operating contexts, including urban-rural robustness checks. The recipient-side willingness-to-pay estimates and a mixed logit robustness check are reported in Supplementary Information. Role gaps for privacy, safety and noise We summarise how strongly each cue is weighted when the same participants evaluate scenarios as bystanders versus as recipients. For each attribute, we computed a role-gap odds ratio (OR) defined as: $$\:Role-gap\:odds\:ratio=\:\frac{ORbystanders}{ORrecipients}$$ This ratio captures how much more (or less) strongly an attribute influences choice under the bystander viewpoint relative to the recipient viewpoint. The numerator is the odds ratio for the attribute computed under the bystander coefficients (recipient baseline plus bystander shift), and the denominator is the odds ratio computed under the recipient baseline alone. We use this ratio rather than raw coefficient differences because it provides a scale-free, multiplicative measure that is directly interpretable. Values greater than 1 indicate the contrast is associated with higher relative odds under the bystander viewpoint than under the recipient viewpoint, while values less than 1 indicate the opposite. Confidence intervals are delta-method intervals based on the model's robust covariance. Table 2 and Fig. 1 present the role-gap odds ratios for the full sample alongside urban and rural subsamples. Supplementary Information in S4 (Supplementary Table S5 & Fig. S1 ) reports the underlying odds ratios separately for each role. To assess whether the role-gap patterns are robust across settlement types, we re-estimated the full interaction model separately for urban (n = 599) and rural (n = 402) respondents and computed role-gap odds ratios from each. Because separate estimation does not constrain the error scale to be equal across subsamples, cross-area comparisons of raw coefficient magnitudes are subject to potential scale differences [ 40 ]. We therefore present the urban–rural analysis as a pattern-replication check rather than a formal magnitude comparison. The role-gap odds ratios reported here (defined as within-model ratios of bystander to recipient odds) are inherently less sensitive to this scale concern than direct cross-model coefficient comparisons. Supplementary Information S5 (Supplementary Table S6 & Fig. S2) reports the corresponding odds ratios by role and area. Table 2 Urban and rural role-gap odds ratio Label Area OR gap Lower 95% CI for OR gap Higher 95% CI for OR gap One unit noisier — Noise-sensitive areas Full sample 0.990 0.950 1.031 Urban 0.966 0.915 1.020 Rural 1.023 0.960 1.090 One unit noisier – Non-sensitive areas Full sample 0.940 0.901 0.981 Urban 0.949 0.896 1.004 Rural 0.926 0.867 0.989 Camera vs No camera — Public area Full sample 1.392 1.192 1.626 Urban 1.515 1.233 1.862 Rural 1.247 0.985 1.577 Camera vs No camera— Private area Full sample 0.741 0.653 0.840 Urban 0.729 0.623 0.853 Rural 0.752 0.609 0.928 Drone registered vs NOT registered Full sample 1.042 0.952 1.141 Urban 0.991 0.881 1.115 Rural 1.120 0.971 1.292 Operator licensed vs NOT licensed Full sample 1.139 1.046 1.240 Urban 1.182 1.055 1.324 Rural 1.102 0.967 1.255 Noise was coded so that a one-unit increase corresponds to one step noisier. In the pooled results, noisier operations reduced acceptability more under the bystander role than under the recipient role in nonsensitive settings. In noise-sensitive settings, this role gap vanished: the two roles converged on similar noise sensitivity. This pattern is consistent with recipients becoming more attentive to noise when the setting signals higher sensitivity to disturbance, narrowing the gap with bystanders. The urban and rural models showed the same direction in nonsensitive settings. In noise-sensitive settings, neither subsample showed a significant role gap: the urban estimate was 0.97 and the rural estimate was 1.02, both with intervals spanning unity. Overall, settlement type did not overturn the main pattern. Camera contrast is reported as camera present versus no camera, separated by public and private delivery points. In public delivery points, the role-gap OR was 1.39, indicating that bystanders weighted camera presence more favourably than recipients; consistent with cameras signalling accountability in shared spaces. In private delivery points, the direction reversed: the role-gap OR fell to 0.74, indicating that bystanders penalised cameras more than recipients near private property. This cross-context reversal indicates that privacy-related cues depend jointly on role and place. The urban and rural subsamples preserved this public–private pattern (urban: 1.52 vs 0.73; rural: 1.25 vs 0.75). Confidence intervals were wider in rural areas and the rural public-delivery-point estimate did not reach statistical significance. For institutional safeguards, the pooled results show a modest licensing role gap, with bystanders placing slightly more weight on licensing than recipients. Registration showed little evidence of a role gap, suggesting broadly similar weighting across roles. The urban and rural subsamples were consistent with this picture: the licensing gap was clearest in the urban model while the rural estimate was less precise. Registration remained near parity in both settings. Taken together, safeguards mattered under both roles, but they did not separate bystanders from recipients as strongly as noise and cameras did. Overall, the role shift changed preferences most for cameras and noise, while safety attributes were more stable across roles. Camera preferences showed a cross-context reversal: bystanders favoured cameras at public delivery points but penalised them near private ones, while recipients were largely indifferent. For noise, bystanders reacted more strongly than recipients in routine settings, but the two roles converged in noise-sensitive places. Licensing and registration were valued under both roles, with only modest role differences. These patterns were broadly consistent across urban and rural subsamples, though rural estimates were less precise for some contrasts. Recipient-side willingness-to-pay estimates, which translate these preference weights into monetary equivalents, are reported in Supplementary Information S7. Discussion In this study, we show that acceptance of drone delivery is not captured by a single public preference. Existing evidence on societal acceptance often reports population-level attitudes or customer-side trade-offs, and it repeatedly identifies noise, privacy and safety as central concerns [ 16 , 41 ]. By asking the same respondents to evaluate matched drone delivery scenarios as recipients and as bystanders, we find that encounter role changes the relative weight placed on these concerns. Bystanders penalised noisy operations more in everyday settings, but this role gap vanished in noise-sensitive locations, where recipients converged on similar noise sensitivity. Camera preferences showed a striking cross-context reversal: bystanders favoured cameras at public delivery points but penalised them near private ones, while recipients were largely indifferent. In contrast, institutional safeguards were valued under both roles, suggesting that licensing and registration operate as broadly shared assurance cues. These role-dependent patterns were broadly consistent across urban and rural respondents. These role gaps are consistent with a straightforward logic about who benefits and who bears the costs of a flight. Research on risk perception shows that judgements about new technologies depend on factors such as perceived lack of control and uncertainty, which can increase sensitivity to potential harms when people feel exposed rather than protected [ 42 ]. Trust in the organisations that manage risk also shapes how people judge hazards, especially when individuals cannot directly monitor operations or intervene [ 24 ]. In our study, the bystander role approximates this low-control position: bystanders cannot opt out of overflight, gain no service benefit, and have no say over routing or timing. The recipient role, by contrast, makes the delivery benefit immediate and personal, which may offset concerns about noise or surveillance. More broadly, work on social acceptance of infrastructure argues that acceptance depends not only on average attitudes but on how local impacts and broader benefits are distributed across affected groups [ 6 ]. When the costs of a technology fall disproportionately on those who do not share in its benefits acceptance among the cost-bearing group may erode even if aggregate sentiment is positive [ 6 ]. For drone delivery, this distributive logic implies that "public acceptance" should be decomposed into at least two linked perspectives within the same people. This calls for service designs and governance frameworks that reduce impacts on non-recipients while making accountability visible to everyone. Noise emerged as a robust constraint in our experiment, but its salience depended on both encounter role and operating context. Noisier flights reduced acceptance more strongly under the bystander viewpoint than under the recipient viewpoint, which fits the idea that external costs carry more weight when the service benefit is absent. When the scenario shifted to noise-sensitive places such as homes and hospitals, the role gap vanished, suggesting that tolerance narrows when delivery is imagined close to private life or vulnerable settings. This context-dependent noise sensitivity aligns with evidence from environmental acoustics showing that annoyance responses to the same sound level depend strongly on the activity being disrupted and the expectation of quiet in a given location [ 43 ]. It is also consistent with public health guidance and epidemiological reviews that treat environmental noise as a material health risk and support targeted mitigation where exposure is most consequential [ 19 , 44 ]. More specifically, psychoacoustic studies of small multirotor drones show that their tonal and high-frequency features generate disproportionate annoyance relative to conventional transport sources at equivalent exposure levels [ 45 , 46 ], which may amplify the context effect we observe. For operators and regulators, the practical implication is to manage noise through both engineering and operations: prioritising lower-noise propulsion designs, reducing hovering and low-altitude manoeuvres near dwellings, and routing away from noise-sensitive land uses. Recipient willingness-to-pay estimates (Supplementary Information S7) reinforce that quieter operations carry tangible monetary value. Camera use produced the clearest evidence that privacy concerns are context-specific and role-dependent. Prior work shows that privacy preferences can change sharply across situations, even for the same person, because privacy judgements depend on what information is collected, by whom, and in which setting [ 47 , 48 ]. This helps interpret our cross-context pattern. Camera use did not act as a uniform privacy penalty. Under the bystander viewpoint, cameras were more acceptable in public delivery points, but were penalised at private delivery points, consistent with stronger privacy expectations near homes and private property. Under the recipient viewpoint, camera effects were comparatively small. Compared with studies that mainly measure customer-side privacy trade-offs for drone delivery, our results suggest that a single privacy effect is not stable across roles or settings. A practical implication is that camera policies should be framed as part of the service context, not as a fixed feature. Operators may reduce conflict by limiting data collection and retention by default (following GDPR), using clear purpose limitation, and adopting context-dependent controls such as geofenced camera restrictions [ 9 ]. In contrast, licensing and registration were valued under both roles. Recipient side evidence show that these safety measures carry the largest utility and monetary values (Supplementary information S7) among the attributes we tested. This pattern suggests that institutional safeguards function as broad assurance cues that support perceived accountability and the governance of risk, especially where individuals cannot directly monitor the system [ 24 ]. The importance of identifiable and enforceable oversight is consistent with legitimacy research, which argues that rule following is closely tied to whether authorities and rules are seen as credible and justified. It also aligns with the direction of aviation governance, where identification tools are used to make operators traceable. For example, Remote ID in the United States and Europe is designed to provide identification and location information during flight and to help authorities respond to unsafe or unauthorised operations [ 49 , 50 ]. While our study focuses on licensing and registration rather than Remote ID, the shared function of these attributes is to make responsibility clear when things go wrong. For implementation, the key is visibility. Operators and regulators can make licensing and registration easy to verify at the point of encounter through markings, simple public-facing identifiers, and app-based notifications. More generally, evidence from regulated sectors suggests that well-framed transparency can support citizen trust, although effects are modest and context-dependent [ 51 ]. This reinforces the value of pairing formal safeguards with clear public communication that signals active oversight. We also examined whether the role-dependent patterns hold across settlement types, because everyday exposure and expectations about flight frequency can differ across urban, suburban and rural environments. Across all domains, the qualitative role-gap patterns replicated: noise and camera cues showed the strongest role and context dependence, while licensing and registration remained valued under both encounter roles. Beyond overall replication, one noteworthy area-level difference emerged for cameras. In public delivery points, the urban role-gap OR was above unity, while the rural estimate was smaller and its confidence interval included unity. This suggests that the bystander camera premium in public spaces may be more salient in denser urban environments where shared-space encounters are frequent, whereas rural respondents show more heterogeneous views. More broadly, several rural estimates had wider confidence intervals, consistent with prior research showing that perceived risks such as noise and privacy are central drivers of attitudes towards drone delivery, and that these concerns can be shaped by context and experience rather than being fixed traits [ 13 , 16 ]. Limitations should be considered when interpreting the magnitudes of our estimates. First, the data come from an online survey in Poland (February to March 2022) with quota targets, which supports internal comparison but does not remove all risks of selection bias that can arise in web surveys [ 52 ]. Second and similarly, as shown in Supplementary Information S1, our sample skews younger than the national population. This may bias responses toward individuals who are more technologically receptive, potentially underrepresenting the perspectives of older and less technology-confident groups. Third, choice tasks are stated preferences. They allow controlled trade-offs, but respondents may still behave differently when real services, prices and repeated exposure are involved [ 53 , 54 ]. Fourth, encounter role is cued through framing. This is central to our design, but it also means that role differences can be influenced by how scenarios are described, and future field work could test whether similar gaps appear when people experience flights in real operating environments. Finally, while a mixed logit robustness check (Supplementary Information S6, supplementary table s7) confirms that the key role-dependent patterns are preserved after accounting for unobserved preference heterogeneity, our models do not explore systematic sources of this heterogeneity. For instance, whether the role gap varies by age, gender, income, or prior familiarity with drones. Future work could use latent class or hybrid choice specifications to identify whether particular population segments show larger or smaller role effects. Separately, the experimental design could be extended to disentangle specific camera functions (e.g., navigation vs recording), data retention policies, and flight frequency, which would help test how stable the patterns are as drone services move toward mass deployment. Despite these caveats, the results point to a clear and actionable conclusion. Social acceptance of drone delivery is not a single attitude. It depends on where the flight occurs and on whether people experience the operation as recipients or as bystanders. This matters because drone delivery will only scale if services earn acceptance not just among customers but also among the wider communities in whose spaces flights take place [ 6 ]. Our findings imply that operators and regulators should treat noise and camera policies as context-sensitive design choices, while using licensing and registration as visible assurance tools that apply across roles. Designing for both encounter roles can strengthen accountability and improve the prospects for drone delivery that is compatible with diverse public expectations. Methods Survey design The questionnaire was designed to elicit preferences for drone delivery operations, with a focus on operational externalities and governance cues that are salient to public acceptance. The instrument combined sociodemographic and attitudinal questions with two stated preference discrete-choice experiments. Survey development followed an iterative process. First, two rounds of consultations with local authorities and other stakeholders were used to identify plausible drone delivery use cases and the operational issues most likely to shape acceptance. These inputs were translated into an initial stated preference experiment presented in a tabular format. Early testing indicated that respondents struggled to visualise the operating context from tables alone. To improve comprehension, we moved to a visual choice format. The visual stimuli were produced by graphic designers based on the attribute combinations specified in the experimental design. A second round of testing suggested that the visual format reduced respondent burden and improved clarity. Items that remained unclear were modified or removed before final deployment. The final survey was administered in Polish. Consent materials were also presented in Polish before the questionnaire began. The survey included two groups of stated preference discrete-choice experiments administered to the same individuals. Each choice task presented two unlabelled drone alternatives (Drone A and Drone B) and an opt-out option (“Neither”). Figures 2 and 3 include examples of a choice task presented to respondents, while the upper level shows the bystander tasks and the lower level presents the recipient tasks. The recipient experiment included service attributes that are only meaningful under the recipient framing (delivery time and delivery cost), while the bystander experiment included flight topography. In the primary pooled model, attributes common to both roles enter with recipient-baseline and bystander-shift parameters. Attributes unique to one role are incorporated through role-gated terms that restrict each to the relevant task block (with values set to zero outside the applicable role). This specification includes all attributes shown to respondents while preserving direct within-person comparability for the shared attributes. For recipient-side willingness-to-pay estimation (Supplementary Information S7, Supplementary Table S8, S9 & Fig. S3), we estimate a separate recipient-only model that includes delivery cost, scaled in 5 PLN units. In addition to the discrete-choice tasks, the questionnaire collected background information to describe the sample and to support subgroup analyses. Respondents reported sociodemographic characteristics (age, gender, household composition, employment status, dwelling type, and settlement type). Settlement type was used to define urban (n = 599) and rural (n = 402) subgroups for robustness checks. The questionnaire also collected measures of everyday delivery practices (delivery frequency, typical delivery locations, and approximate monthly spending on delivery services); these variables are not used in the choice models but provide descriptive context reported in Supplementary Information S1. Finally, we measured familiarity with drones (whether respondents had previously seen a drone) and attitudes toward drone delivery. The recipient block always preceded the bystander block, so encounter role is partially confounded with block order. The within-person design still enables direct comparison of role-framed valuations within respondents, but learning, anchoring, or fatigue may also contribute to the observed recipient–bystander differences. We therefore interpret the estimates as role-framed within-person contrasts and note that future experiments should randomise block order to isolate pure role effects. Data collection Data were collected in Upper Silesian Region (GZM) in Poland in February–March 2022 using an online questionnaire. A professional survey company was responsible for recruiting participants based on sample requirements (age, gender, and type of living area) to approximate census benchmarks for the study area. The survey remained in the field for approximately two weeks and the mean completion time was about 6 minutes, which is consistent with the task structure (13 choice tasks plus short background sections) and the visual format that simplified scenario comprehension. The analytical sample comprises 1,001 respondents who provided usable responses for the discrete-choice tasks. The recruitment contract specified delivery of completed questionnaires; accordingly, the dataset contains no incomplete choice tasks. The study received ethics approval from University College London (UCL) (Project Ethics Identification: 20751/001). All respondents provided informed consent prior to participation, and were compensated through the panel provider’s standard incentive scheme. During fieldwork, data were stored on an EU-based cloud server hosted by a project partner; after survey completion, the data were transferred to UCL and stored in the project’s shared repository for analysis. The dataset contains no identifiable personal information. Data preparation and analysis The bystander and recipient choice files were linked by respondent identifier and combined into a single pooled dataset. The choice outcome was coded as a three-category variable (Drone A, Drone B, or Neither). Binary attributes were coded as one-sided indicators with clear baselines: operator licensing (1 = not licensed, 0 = licensed), drone registration (1 = not registered, 0 = registered), and camera (1 = no camera, 0 = camera present). Noise was treated as an ordered five-level variable oriented so that a one-unit increase corresponds to one step noisier. Built context was recoded to a binary indicator (1 = noise-sensitive: home or hospital; 0 = other), and delivery point to a binary indicator (1 = private: doorstep, terrace, window; 0 = public/communal). Parcel type was recoded into three categories (medical, food, other as baseline). For recipient tasks, delivery cost was entered in 5 zł units. Role-gated variables (cost, time for recipients; topography dummies for bystanders) were set to zero outside the applicable role block to avoid missing-value propagation. Our primary goal was to estimate the within-person change in attribute sensitivity between encounter roles. We analysed the pooled choice data using a panel multinomial logit model with three alternatives per task. Because each respondent completed repeated tasks across both role blocks, we treated the data as panel data and used respondent-clustered (panel-robust) standard errors. To quantify how encounter role reweights the same attributes, we used a role indicator r_nt coded as 0 for recipient tasks and 1 for bystander tasks. For shared attributes, coefficients under the recipient role serve as the baseline; bystander coefficients are expressed as the baseline plus an additive bystander shift. Attributes unique to one role enter through role-gated terms: delivery cost and time are multiplied by (1 – r nt ) so they contribute only in recipient tasks, while topography dummies are multiplied by r_nt so they contribute only in bystander tasks. This parameterisation supports direct within-sample tests of role differences for each attribute and interaction term. Specifically, let respondent n = 1,…, N evaluate task t = 1,…,T nt over three alternatives j∈{A,B,N} (where N denotes “Neither”). Let r nt ∈{0,1} be the role indicator (0 = recipient viewpoint, 1 = bystander viewpoint). This parameterisation means coefficients under the bystander viewpoint are the sum of the recipient baseline and the bystander shift. Utilities are alternative-specific and linear in attributes. We write utility as \(\:{U}_{ntj}=\:{V}_{ntj}+\:{\epsilon\:}_{ntj}\) , where \(\:{\epsilon\:}_{ntj}\:\) representing type I extreme value. Under the i.i.d. Type I extreme value assumption, the probability that respondent n chooses alternative j in task t is: We estimate a pooled multinomial logit with role-specific slopes and context interactions. The role indicator r n equals 1 when the same participant evaluates scenarios as a bystander and 0 when evaluating as a recipient (two viewpoints on one sample). The vector of drone attributes for alternative j in task t comprises: Noise ntj – Ordered noise score, coded so that a one-unit increase is one step noisier (louder); Sense ntj – built-context indicator (1 = noise-sensitive: home or hospital; 0 = other); Lic ntj ∈{0,1} – licensing coding (1 = not licensed, 0 = licensed); Reg ntj ∈{0,1} – registration coding (1 = not registered, 0 = registered); Cam ntj ∈{0,1} – camera coding used in estimation (1 = no camera, 0 = camera present); Priv ntj ∈{0,1} – delivery point indicator (1 = private: doorstep, terrace, window; 0 = public/communal); \(\:\alpha\:\) N,rec – opt-out intercept under the recipient viewpoint; \(\:\alpha\:\) N,by – additive opt-out shift under the bystander viewpoint; In addition, the following role-gated attributes enter utility only for the relevant role: Cost ntj — delivery cost in 5 zł units (recipient tasks only; set to zero for bystander tasks); Time ntj — delivery time (recipient tasks only; set to zero for bystander tasks); Village ntj ∈ {0,1} — topography indicator for over-village flight (bystander tasks only; set to zero for recipient tasks; baseline = over-city); Rural ntj ∈ {0,1} — topography indicator for rural-area flight (bystander tasks only; set to zero for recipient tasks; baseline = over-city); Parcel _med ntj ∈ {0,1} — parcel content indicator for medical deliveries (both roles; baseline = other); Parcel _food ntj ∈ {0,1} — parcel content indicator for food deliveries (both roles; baseline = other). Noise is entered linearly as an ordered 1–5 scale, which imposes a constant marginal effect per step increase in noisiness; we use this parsimonious specification because the stimulus was designed as a uniform step scale. For camera, operator licensing, and drone registration, the model uses "undesirable-state" indicator variables (Cam = 1 when no camera is present, Lic = 1 when not licensed, Reg = 1 when not registered). In Results, we report contrasts oriented toward the desirable state (e.g., "camera present vs no camera," "licensed vs not licensed"). These correspond to a one-unit decrease in the respective indicator (i.e., moving from the undesirable to the desirable state), so the reported odds ratios equal exp(−β) for each attribute. Because Drone A and Drone B are unlabelled (generic) alternatives, we normalise their alternative-specific constants to zero, only the opt-out ("Neither") receives an estimated constant (with a recipient baseline and a bystander shift). This is standard for unlabelled designs where the alternatives carry no inherent identity beyond their attributes [ 35 ]. We verified that introducing separate left/right position constants did not materially change the results. The urban/rural estimation follows the same model structure, estimated separately within urban (n = 599) and rural (n = 402) subsamples. Because separate estimation does not constrain the error scale to be equal across subsamples, we interpret cross-area comparisons as pattern-replication checks rather than formal magnitude tests. To assess whether the pooled MNL results depend on the assumption of homogeneous preferences, we re-estimated the pooled role specification as a mixed logit with normally distributed random coefficients on all four core shared attributes (noise, camera, licensing, and registration). Bystander shifts and context interactions were kept as fixed parameters. The model was estimated by maximum simulated likelihood with 500 Halton draws and panel likelihood across repeated tasks. We report the mixed logit as a robustness check in Supplementary Information S6. The MNL is retained as the primary specification because it provides a more transparent baseline for interpreting role and context interactions, and the mixed logit confirms that the substantive role-gap patterns are preserved after accounting for unobserved heterogeneity. Declarations Data availability The datasets generated and/or analysed during the current study are not publicly available due to ethics restrictions, but are available from the corresponding author on reasonable request. Code availability The R scripts used for computing the results in this manuscript can be found in below github redispository (https://github.com/Basilcao/Drone-manuscript). Acknowledgements The research reported in this paper has been supported by European Union’s Horizon 2020 research and innovation programme under Grant Agreement No 815269, project HARMONY. The contribution of Luciano Pana Tronca is acknowledged for working with GZM for the workshops that fed with insights the questionnaire design. Author contributions MK perceived the idea of this survey. MK, FS and YZ performed study design, data curation and preparation; YC, MK, FS and YZ analyzed and interpreted the data. YC collaborate with the graphic designers for the visualisation of the SPs, prepared all the statistical figures and was a major contributor in writing the manuscript. MK, FS and YZ supervised, reviewed and revised. All author wrote, read, and approved the manuscript. Competing interests The author declare no competing interests. References J.K. Stolaroff, C. Samaras, E.R. O’Neill, A. Lubers, A.S. Mitchell, D. Ceperley, Energy use and life cycle greenhouse gas emissions of drones for commercial package delivery, Nature communications 9(1) (2018) 409. A. Raghunatha, P. Hilletofth, P. Thollander, Exploring the dynamics of public acceptance factors and drone transport planning towards sustainable transitions, European Transport Research Review 17(1) (2025) 1-14. V. Garg, S. Niranjan, V. Prybutok, T. Pohlen, D. Gligor, Drones in last-mile delivery: A systematic review on Efficiency, Accessibility, and Sustainability, Transportation Research Part D: Transport and Environment 123 (2023) 103831. A. Zenz, J. Powles, Resisting technological inevitability: Google Wing’s delivery drones and the fight for our skies, Philosophical Transactions A 382(2285) (2024) 20240107. S. Zhang, P. Anciaes, E. Chaniotakis, The future of urban deliveries? Perceptions and intentions about delivery robots and drones in Europe, Springer2025. R. Wüstenhagen, M. Wolsink, M.J. Bürer, Social acceptance of renewable energy innovation: An introduction to the concept, Energy policy 35(5) (2007) 2683-2691. C. Al Haddad, E. Chaniotakis, A. Straubinger, K. Plötner, C. Antoniou, Factors affecting the adoption and use of urban air mobility, Transportation research part A: policy and practice 132 (2020) 696-712. Q. Long, J. Ma, F. Jiang, C.J. Webster, Demand analysis in urban air mobility: A literature review, Journal of Air Transport Management 112 (2023) 102436. N. EASA, Study on the societal acceptance of urban air mobility in Europe, European Union Aviation Safety Agency and McKinsey & Company (2021). S. Hess, K. Train, Correlation and scale in mixed logit models, Journal of choice modelling 23 (2017) 1-8. E. Lancsar, J. Louviere, Deleting ‘irrational’responses from discrete choice experiments: a case of investigating or imposing preferences?, Health economics 15(8) (2006) 797-811. B.I. Scott, K.I. Andritsos, A drone strategy 2.0 for a smart and sustainable unmanned aircraft eco-system in Europe, Air and Space Law 48(3) (2023). W. Yoo, E. Yu, J. Jung, Drone delivery: Factors affecting the public’s attitude and intention to adopt, Telematics and Informatics 35(6) (2018) 1687-1700. N. Wang, N. Mutzner, K. Blanchet, Societal acceptance of urban drones: A scoping literature review, Technology in Society 75 (2023) 102377. R. Khan, S. Tausif, A. Javed Malik, Consumer acceptance of delivery drones in urban areas, International Journal of Consumer Studies 43(1) (2019) 87-101. R. Kellermann, T. Biehle, H. Mostofi, Modelling public attitude towards drone delivery in Germany, European Transport Research Review 15(1) (2023) 38. A.W. Christian, R. Cabell, Initial investigation into the psychoacoustic properties of small unmanned aerial system noise, 23rd AIAA/CEAS aeroacoustics conference, 2017, p. 4051. S.H. Kim, Choice model based analysis of consumer preference for drone delivery service, Journal of Air Transport Management 84 (2020) 101785. WHO, Environmental noise guidelines for European region, World Health Organization, 2019. P.J. Lee, C. Hopkins, R. Penedo, Attitudes to noise inside dwellings in three megacities: Seoul, London, and São Paulo, International Journal of Environmental Research and Public Health 17(16) (2020) 6005. S. Leon, C. Chen, A. Ratcliffe, Consumers’ perceptions of last mile drone delivery, International Journal of Logistics Research and Applications 26(3) (2023) 345-364. R. Altawy, A.M. Youssef, Security, privacy, and safety aspects of civilian drones: A survey, ACM Transactions on Cyber-Physical Systems 1(2) (2016) 1-25. Y. Wang, H. Xia, Y. Huang, Examining American and Chinese internet users' contextual privacy preferences of behavioral advertising, Proceedings of the 19th ACM conference on computer-supported cooperative work & social computing, 2016, pp. 539-552. M. Siegrist, Trust and risk perception: A critical review of the literature, Risk analysis 41(3) (2021) 480-490. S.A. Applin, Deliveries by drone: Obstacles and sociability, The Future of Drone Use: Opportunities and Threats from Ethical and Legal Perspectives, Springer2016, pp. 71-91. A. AlHogail, Improving IoT technology adoption through improving consumer trust, Technologies 6(3) (2018) 64. E. Öz, E. Heikkilä, R. Tiusanen, Development of an Organisational Certification Process for Specific Category Drone Operations, Drones 6(10) (2022) 278. Y. Zhang, M. Kamargianni, A review on the factors influencing the adoption of new mobility technologies and services: autonomous vehicle, drone, micromobility and mobility as a service, Transport reviews 43(3) (2023) 407-429. S.N. Lingam, R. Verstegen, S.M. Petermeijer, M. Martens, Human interactions with delivery drones in public spaces: design recommendations from recipient and bystander perspectives, Frontiers in Robotics and AI 12 (2025) 1580289. A. Straubinger, U. Kluge, M. Fu, C. Al Haddad, K.O. Ploetner, C. Antoniou, Identifying demand and acceptance drivers for user friendly urban air mobility introduction, Towards User-Centric Transport in Europe 2: Enablers of Inclusive, Seamless and Sustainable Mobility, Springer2020, pp. 117-134. Y.-J. Tu, S. Piramuthu, Security and privacy risks in drone-based last mile delivery, European Journal of Information Systems 33(5) (2024) 617-630. Q. Tan, H. Bian, J. Guo, P. Zhou, H.K. Lo, S. Zhong, X. Zhang, Virtual flight simulation of delivery drone noise in the urban residential community, Transportation Research Part D: Transport and Environment 118 (2023) 103686. L.K.L. Tan, B.C. Lim, G. Park, K.H. Low, V.C.S. Yeo, Public acceptance of drone applications in a highly urbanized environment, Technology in Society 64 (2021) 101462. L.Y. Koh, J.Y. Lee, X. Wang, K.F. Yuen, Urban drone adoption: Addressing technological, privacy and task–technology fit concerns, Technology in Society 72 (2023) 102203. J.J. Louviere, D.A. Hensher, J.D. Swait, Stated choice methods: analysis and applications, Cambridge university press2000. M. Kamargianni, S. Dubey, A. Polydoropoulou, C. Bhat, Investigating the subjective and objective factors influencing teenagers’ school travel mode choice–An integrated choice and latent variable model, Transportation Research Part A: Policy and Practice 78 (2015) 473-488. S. Hess, D. Palma, Apollo: A flexible, powerful and customisable freeware package for choice model estimation and application, Journal of choice modelling 32 (2019) 100170. T. Islam, J.J. Louviere, P.F. Burke, Modeling the effects of including/excluding attributes in choice experiments on systematic and random components, International Journal of Research in Marketing 24(4) (2007) 289-300. D.G. Fiebig, M.P. Keane, J. Louviere, N. Wasi, The generalized multinomial logit model: accounting for scale and coefficient heterogeneity, Marketing science 29(3) (2010) 393-421. J. Swait, J. Louviere, The role of the scale parameter in the estimation and comparison of multinomial logit models, Journal of marketing research 30(3) (1993) 305-314. A. Berke, G. Ding, C. Chin, K. Gopalakrishnan, K. Larson, H. Balakrishnan, M.Z. Li, Drone delivery and the value of customer privacy: A discrete choice experiment with US consumers, Transportation Research Part C: Emerging Technologies 157 (2023) 104391. P. Slovic, Perception of bisk, Emerging Technologies, Routledge2020, pp. 141-146. R. Guski, D. Schreckenberg, R. Schuemer, WHO environmental noise guidelines for the European region: A systematic review on environmental noise and annoyance, International journal of environmental research and public health 14(12) (2017) 1539. T. Münzel, F.P. Schmidt, S. Steven, J. Herzog, A. Daiber, M. Sørensen, Environmental noise and the cardiovascular system, Journal of the American College of Cardiology 71(6) (2018) 688-697. C. Kawai, J. Jäggi, F. Georgiou, J. Meister, R. Pieren, B. Schäffer, Short-term noise annoyance towards drones and other transportation noise sources: A laboratory study, The Journal of the Acoustical Society of America 156(4) (2024) 2578-2595. N.S. Zawodny, A. Christian, R. Cabell, A summary of NASA research exploring the acoustics of small unmanned aerial systems, 2018 AHS Technical Meeting on Aeromechanics Design for Transformative Vertical Flight, 2018. A. Acquisti, L. Brandimarte, G. Loewenstein, Privacy and human behavior in the age of information, Science 347(6221) (2015) 509-514. H. Nissenbaum, Privacy as contextual integrity, Wash. L. Rev. 79 (2004) 119. E. Commission, Commission Implementing Regulation (EU) 2019/947 of 24 May 2019 on the rules and procedures for the operation of unmanned aircraft, Official Journal of the European Union L 152, EU, 2019. U.S.G.A. Office, Drones: Actions needed to better support remote identification in the national airspace (GAO-24-106158), in: GAO (Ed.) Washington, DC, 2024. S. Grimmelikhuijsen, F. de Vries, R. Bouwman, Regulators as guardians of trust? The contingent and modest positive effect of targeted transparency on citizen trust in regulated sectors, Journal of Public Administration Research and Theory 34(1) (2024) 136-149. J. Bethlehem, Selection bias in web surveys, International statistical review 78(2) (2010) 161-188. R.T. Carson, T. Groves, Incentive and informational properties of preference questions, Environmental and resource economics 37(1) (2007) 181-210. J.J. Murphy, P.G. Allen, T.H. Stevens, D. Weatherhead, A meta-analysis of hypothetical bias in stated preference valuation, Environmental and Resource Economics 30(3) (2005) 313-325. Additional Declarations No competing interests reported. Supplementary Files SupplementaryInformationrevised.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 01 May, 2026 Reviewers agreed at journal 01 May, 2026 Reviewers agreed at journal 30 Apr, 2026 Reviewers invited by journal 24 Apr, 2026 Editor assigned by journal 24 Apr, 2026 Submission checks completed at journal 23 Apr, 2026 First submitted to journal 14 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9412457","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":634379810,"identity":"2a60c934-192f-4b5a-95aa-9bc9774c1fdf","order_by":0,"name":"Maria Kamargianni","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDElEQVRIiWNgGAWjYFACNiQqoQJEMjcwk6DlDIhkJFILGDC2EaFFt4Et8dONmnt2fexnjz14OG+bPH//wcbPBQw2+fIO2LWYHWA7LJ1zrDi5jScv3SBx223DGTcSm6VnMKRZbjyASwt7g3QOW0IyG0OOmQRQC2PDDcYGaR6GwwaGDTi1NP/O+QfUwv8GqGXObfv55w82/8avhe2YdG5bgh2bBMiWhtuJGw4ktoFtkcfhfbPDbGnWuX0JCWwSQFsSjt1O3ngjsc16hkGagQEuLcfbjG/nfEuwl+/PMZP8UXPbdt75w4dvF1TYGMjjcBgDNAoS0eSBVhgcwKEFCuwxhXDaMgpGwSgYBSMNAADCuFwGbg0CwwAAAABJRU5ErkJggg==","orcid":"","institution":"University College London","correspondingAuthor":true,"prefix":"","firstName":"Maria","middleName":"","lastName":"Kamargianni","suffix":""},{"id":634379811,"identity":"7b5c9098-e86e-424d-8e9a-7a5537040e6b","order_by":1,"name":"Yuerong Zhang","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Yuerong","middleName":"","lastName":"Zhang","suffix":""},{"id":634379812,"identity":"33ad36f0-2d96-4bb7-858a-acfeb659c65f","order_by":2,"name":"Fangqing Song","email":"","orcid":"","institution":"University of Leeds","correspondingAuthor":false,"prefix":"","firstName":"Fangqing","middleName":"","lastName":"Song","suffix":""},{"id":634379813,"identity":"1f3f9038-ad71-48bd-bec8-34b0b6f3ae3c","order_by":3,"name":"Yanghui Cao","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Yanghui","middleName":"","lastName":"Cao","suffix":""}],"badges":[],"createdAt":"2026-04-14 08:23:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9412457/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9412457/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108804064,"identity":"9f42aceb-1718-4206-a746-98e3aa61d386","added_by":"auto","created_at":"2026-05-08 15:15:14","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":114016,"visible":true,"origin":"","legend":"\u003cp\u003eCombined role-gap odds ratio\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9412457/v1/559fbf397c648b76ca2e19f5.png"},{"id":108502584,"identity":"56999991-5974-4a3a-b2cb-2b9ebe796f3b","added_by":"auto","created_at":"2026-05-05 11:03:54","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":331427,"visible":true,"origin":"","legend":"\u003cp\u003eExamples of the choice tasks (recipient)\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9412457/v1/bd0d10d928760e4204cfda0f.png"},{"id":108804527,"identity":"01a927b6-4617-4428-a096-7edf9dd3ee03","added_by":"auto","created_at":"2026-05-08 15:21:20","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":387226,"visible":true,"origin":"","legend":"\u003cp\u003eExamples of the choice tasks (bystander)\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9412457/v1/10cc3e0512179630aaf2f1ff.png"},{"id":108811705,"identity":"576db3f5-d8d5-4a1f-ad2e-bf2d41249fa5","added_by":"auto","created_at":"2026-05-08 16:06:50","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1217304,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9412457/v1/5def4452-9997-43cd-b2fb-f416434c7983.pdf"},{"id":108502582,"identity":"dc06d856-d7b0-4d47-9aaf-199fcde4f517","added_by":"auto","created_at":"2026-05-05 11:03:54","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":467556,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformationrevised.docx","url":"https://assets-eu.researchsquare.com/files/rs-9412457/v1/d094aba32f46f4b1c7087455.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Bystander and recipient roles reweight noise, privacy, and safety in drone deliveries","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDrone delivery is increasingly proposed as a lower‑emission option for the last mile, particularly for small, time‑sensitive payloads. Life‑cycle analysis shows that, when carefully deployed (e.g., light packages, low‑carbon electricity), drone delivery can reduce energy use and greenhouse‑gas emissions relative to diesel trucks [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Especially in favourable scenarios for 0.5 kg packages, the corrected estimate can reduce up to ~\u0026thinsp;54% GHGs than truck delivery [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Beyond emissions, drone delivery may improve service accessibility for populations that face barriers to conventional logistics. For instance, people with limited mobility or those living in areas poorly served by road-based transport [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. However, operational benefits alone do not guarantee viable services. Drone delivery may only scale if operations are accepted in the places where flights occur, including acceptance by people who do not receive the drone deliveries [\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eExisting research on drone delivery acceptance has largely treated the public as a single group. Most studies ask respondents to evaluate scenarios exclusively as prospective users of the service (see, for example, [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]). Large official surveys also report acceptance at the level of citizens as a whole, focusing on noise, privacy, and safety without separating groups by encounter role [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Importantly, these roles are not fixed segments of the population: the same individual may be a recipient in one encounter and a bystander in the next. Our within-person design leverages this by asking each respondent to evaluate scenarios under both perspectives, so that role differences are identified within individuals rather than between groups. To our knowledge, no study has asked the same individuals to evaluate matched drone-delivery scenarios from both the recipient and bystander perspective and then quantified how this role shift changes attribute valuations. Within-person designs are important here because they isolate role effects from stable individual differences (e.g., risk aversion, technology affinity) that would confound between-group comparisons [\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The idea that the same technology is judged differently by beneficiaries and cost-bearers has been explored in other infrastructure contexts. For example, in the transport domain, W\u0026uuml;stenhagen et al. [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]distinguish community acceptance from market acceptance precisely because affected non-users may weight impacts differently from customers. This gap limits the practical value of existing evidence: without knowing whether bystanders and recipients weight the same attributes differently, policymakers and operators lack the basis for role-sensitive service design.\u003c/p\u003e \u003cp\u003eDeployment in everyday spaces confronts three salient externalities: noise, privacy intrusion, and safety risk [\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. For noise, laboratory and field evidence shows that drone sounds are perceived as substantially more annoying than comparable ground-vehicle or conventional aircraft noise at equivalent exposure levels, creating a psychoacoustic burden distinct from familiar transport sources [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Public health guidance stresses context-specific management of noise exposure, recognising it as a material risk factor for wellbeing [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. For privacy, concerns centre on the perceived surveillance potential of onboard cameras, which citizens judge differently depending on whether a drone operates near private property or in a public space [\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. From the broader citizen perspective, privacy and security are consistently identified as central barriers to acceptance [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. For safety, concerns relate to the credibility of operational oversight: whether operators are identifiable, drones are traceable, and regulatory compliance is enforceable [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Risk research links such governance signals to higher trust and lower perceived risk, especially for novel technologies and for observers who lack direct control [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. In the drone context, studies have advocated registration and remote identification to enable accountability and enhance public trust [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. These three dimensions map directly onto policy levers \u0026mdash; noise routing, camera and data-use regulation, and licensing regimes; making it essential to understand how different people value each one [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. We therefore treat noise, privacy, and institutional safeguards as a minimal set of attributes that capture how citizens experience drone operations in daily life, and examine how encounter role shapes their relative influence.\u003c/p\u003e \u003cp\u003eBeyond the role distinction, the operational context of a drone flight is likely to moderate the perception of externalities. For noise, public-health guidance stresses context-specific exposure management, implying that sensitivity should be heightened near homes and hospitals; places where quiet is both expected and health-relevant [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Psychoacoustic evidence reinforces this idea that drone sounds are judged more annoying than comparable transport noise at equal exposure [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], and this penalty should be most salient where ambient noise is lowest and expectations of quiet are strongest *e.g., hospitals). For privacy, the evaluation of an onboard camera is likely to depend on whether the delivery occurs at a public location (e.g., a park or commercial area) or a private one (e.g., a residential doorstep), since privacy norms differ sharply across these settings [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. We therefore model interactions between role and these place-level variables. In addition, we examine urban\u0026ndash;rural differences as a robustness check, based on the self-reported living area from respondents. Urban and rural residents differ in baseline ambient noise, housing density, and proximity to potential flight paths, all of which may moderate externality perceptions [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Prior work has examined drone acceptance across geographical contexts but has predominantly focused on urban populations [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], whereas comparative urban\u0026ndash;rural evidence remains scarce despite the considerable potential of drone delivery in underserved rural areas. We test whether the role-dependent patterns observed in the pooled sample replicate across urban and rural subsamples, strengthening the generalisability of the role-based account.\u003c/p\u003e \u003cp\u003eTo test these expectations, we designed a within-person discrete-choice experiment fielded in Poland (n\u0026thinsp;=\u0026thinsp;1,001). This design follows established practice in stated‑choice experiments and links to standard logit foundations used for behavioural choice modelling [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Each respondent evaluated two blocks of hypothetical drone-delivery scenarios (i.e. one from the bystander perspective and one from the recipient perspective), enabling direct comparison of how the same individual values the same attributes under each role. We estimate pooled panel multinomial logit models in which a role indicator is interacted with each attribute, so that a single model yields separate coefficient estimates for bystanders and recipients. In plain terms, the interaction structure lets us ask: does switching from recipient to bystander change how much a person cares about noise, cameras, licensing, or registration? The experiment included additional scenario attributes that differed across blocks: recipients saw delivery cost and time, while bystanders saw flight topography. These role-specific attributes enter the model through role-gated terms that restrict their contribution to the relevant subset of tasks, so that all attributes shown to respondents are included in the primary specification. A mixed logit robustness check and urban\u0026ndash;rural subgroup analyses are reported in the Supplementary Information. We use the R package Apollo for estimation and post‑estimation checks [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eUsing this design, we find that encounter role systematically reshapes how people evaluate drone-delivery services. Bystanders penalise noisy operations more strongly than recipients in routine locations (e.g., commercial areas), but this role gap vanishes in noise-sensitive settings such as homes and hospitals, where both roles converge on similar noise sensitivity. Camera preferences are strongly context-dependent: bystanders favour camera-equipped drones at public delivery points but resist them near private ones, whereas recipients show weak camera sensitivity across both contexts. Both roles value operator licensing and drone registration, with bystanders placing modestly greater emphasis on compliance. These role-dependent patterns are broadly consistent across urban and rural subsamples.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eWe collected data in the Upper Silesian Region (GZM) in Poland during February and March 2022, yielding 1,001 valid responses with 13,013 evaluated choice tasks. Each respondent completed two blocks of discrete‑choice questions: one block consisting of six scenarios framed as a bystander exposed to a drone delivery, and one block consisting of seven scenarios framed as a recipient of a drone delivery. The survey took place in February\u0026ndash;March 2022. A survey company was responsible for recruiting the participants in our survey based on specific sample requirements (age, gender, and type of living area based on the census data of the GZM Metropolitan area) to secure representativeness of our sample. Sample characteristics and patterns of daily postal practice are reported in Supplementary Information S1 (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e and S2). We estimated pooled panel multinomial logit models with three alternatives (1: drone service A, 2: drone service B, 3: Neither), pooling the two role blocks into a single panel. The attributes and their descriptions shown to respondents are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Attributes were coded as described in Supplementary Information S2 (please see Supplementary Table S3): quietness/noisiness is treated as a five‑icon ordered scale (per‑step increase in noisiness), on‑board camera is coded dummy coded as \"Camera vs No camera\", and operator licensing and drone registration are coded as binary safeguards using \"Licensed vs Not licensed\" and \"Registered vs Not registered\" indicators. Parcel content was recoded into medicine, food, and other (baseline). Context variables capture noise‑sensitive locations (homes/hospitals vs other areas) and private versus public delivery points. Recipient-only attributes (delivery cost and delivery time) and a bystander-only attribute (flight topography) enter through role-gated terms that restrict each to the relevant task block.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eList of attributes and levels\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eParcel type\u003c/b\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAttribute levels (bystander)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAttribute levels (recipient)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedicines, food, clothes, mail, illegal items, surveillance, organic material\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMedicine, food, clothes, mail, surveillance, illegal items\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOperator licensing\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWith license, without license\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWith license, without license\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDrone registration\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRegistered, not registered\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRegistered, not registered\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDrone camera\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWith camera, without camera\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWith camera, without camera\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDrone noise\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 levels (0\u0026ndash;4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5 levels (0\u0026ndash;4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDelivery building\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHospital, home, business/ office area, shopping area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHospital, home, business/ office area, shopping area\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDelivery point\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWindow, doorstep, communal space (e.g., garden), private terrace, open public space\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWindow, doorstep, communal space (e.g., garden), private terrace, open public space\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDelivery cost\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFree, 10 zł, 20 zł, 30 zł, 40 zł, 50 zł\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDelivery time\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20mins, 30mins, 40mins, 50mins, 60mins, 120mins, 6h, 12h, 24h, 48h\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTopography\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOver city, over village, rural area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe primary model includes all attributes shown to respondents. Attributes common to both role blocks (i.e. noise, on-board camera, operator licensing, drone registration, and parcel content) enter utility with a recipient-baseline coefficient and an additive bystander-shift parameter, so that role differences are identified within individuals. Attributes that appeared in only one block (i.e. delivery cost and time for recipients, flight topography for bystanders) are incorporated through role-gated terms, so they are set to non-available for outside the relevant role. This specification preserves direct within-person comparability for the shared attributes [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe primary model specification includes role‑specific interactions: a binary indicator (bystander\u0026thinsp;=\u0026thinsp;1, recipient\u0026thinsp;=\u0026thinsp;0) shifts the slopes of all shared attributes and their context interactions so that differences between viewpoints are identified within individuals. In plain terms, this tests whether the same person values noise, cameras, or safeguards differently when imagining themselves as a bystander versus a recipient. A likelihood‑ratio test strongly favours this interaction model over a pooled main‑effects model without role interactions (ΔLL\u0026thinsp;=\u0026thinsp;58.43; LR\u0026thinsp;=\u0026thinsp;116.86 on 9 df, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Information criteria also favour the interaction model (AIC\u0026thinsp;=\u0026thinsp;25,843.9 vs 25,942.8; BIC\u0026thinsp;=\u0026thinsp;26,008.3 vs 26,039.9), and pseudo‑R\u0026sup2; against equal shares rises from 0.094 to 0.098. Full coefficient tables and fit indices are provided in Supplementary information S3 (Supplementary Table S4). Because the recipient and bystander blocks may differ in random utility scale, we have also estimated an alternative pooled specification that allows a role-specific scale factor for bystander tasks, normalising the recipient scale to one [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. The estimated bystander scale parameter was not significantly different from unity, and the log-likelihood improvement was negligible, indicating that a common-scale specification is adequate (see Supplementary Information S3 for details). The scale factor is not separately identified once role-specific slope shifts are included, so we report the common-scale specification as our main model.\u003c/p\u003e \u003cp\u003eThe following sub-sections report role-gap odds ratios for noise, cameras, and institutional safeguards across operating contexts, including urban-rural robustness checks. The recipient-side willingness-to-pay estimates and a mixed logit robustness check are reported in Supplementary Information.\u003c/p\u003e \u003cp\u003eRole gaps for privacy, safety and noise\u003c/p\u003e \u003cp\u003eWe summarise how strongly each cue is weighted when the same participants evaluate scenarios as bystanders versus as recipients. For each attribute, we computed a role-gap odds ratio (OR) defined as:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Role-gap\\:odds\\:ratio=\\:\\frac{ORbystanders}{ORrecipients}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThis ratio captures how much more (or less) strongly an attribute influences choice under the bystander viewpoint relative to the recipient viewpoint. The numerator is the odds ratio for the attribute computed under the bystander coefficients (recipient baseline plus bystander shift), and the denominator is the odds ratio computed under the recipient baseline alone. We use this ratio rather than raw coefficient differences because it provides a scale-free, multiplicative measure that is directly interpretable. Values greater than 1 indicate the contrast is associated with higher relative odds under the bystander viewpoint than under the recipient viewpoint, while values less than 1 indicate the opposite. Confidence intervals are delta-method intervals based on the model's robust covariance. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e present the role-gap odds ratios for the full sample alongside urban and rural subsamples. Supplementary Information in S4 (Supplementary Table S5 \u0026amp; Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) reports the underlying odds ratios separately for each role.\u003c/p\u003e \u003cp\u003eTo assess whether the role-gap patterns are robust across settlement types, we re-estimated the full interaction model separately for urban (n\u0026thinsp;=\u0026thinsp;599) and rural (n\u0026thinsp;=\u0026thinsp;402) respondents and computed role-gap odds ratios from each. Because separate estimation does not constrain the error scale to be equal across subsamples, cross-area comparisons of raw coefficient magnitudes are subject to potential scale differences [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. We therefore present the urban\u0026ndash;rural analysis as a pattern-replication check rather than a formal magnitude comparison. The role-gap odds ratios reported here (defined as within-model ratios of bystander to recipient odds) are inherently less sensitive to this scale concern than direct cross-model coefficient comparisons. Supplementary Information S5 (Supplementary Table S6 \u0026amp; Fig. S2) reports the corresponding odds ratios by role and area.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUrban and rural role-gap odds ratio\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLabel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOR gap\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLower 95% CI for OR gap\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHigher 95% CI for OR gap\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOne unit noisier \u0026mdash; Noise-sensitive areas\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eFull sample\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.990\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.950\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.031\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eUrban\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.966\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.915\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.020\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRural\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.960\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.090\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOne unit noisier \u0026ndash; Non-sensitive areas\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eFull sample\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.940\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.901\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.981\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eUrban\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.949\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.896\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.004\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRural\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.867\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.989\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCamera vs No camera \u0026mdash; Public area\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eFull sample\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.392\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.192\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.626\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eUrban\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.515\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.233\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.862\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRural\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.247\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.985\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.577\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCamera vs No camera\u0026mdash; Private area\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eFull sample\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.653\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.840\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eUrban\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.729\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.623\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.853\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRural\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.752\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.609\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.928\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDrone registered vs NOT registered\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eFull sample\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.042\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.952\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.141\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eUrban\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.991\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.881\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.115\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRural\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.971\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.292\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOperator licensed vs NOT licensed\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eFull sample\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.139\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.046\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.240\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eUrban\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.182\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.055\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.324\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRural\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.967\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.255\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNoise was coded so that a one-unit increase corresponds to one step noisier. In the pooled results, noisier operations reduced acceptability more under the bystander role than under the recipient role in nonsensitive settings. In noise-sensitive settings, this role gap vanished: the two roles converged on similar noise sensitivity. This pattern is consistent with recipients becoming more attentive to noise when the setting signals higher sensitivity to disturbance, narrowing the gap with bystanders. The urban and rural models showed the same direction in nonsensitive settings. In noise-sensitive settings, neither subsample showed a significant role gap: the urban estimate was 0.97 and the rural estimate was 1.02, both with intervals spanning unity. Overall, settlement type did not overturn the main pattern.\u003c/p\u003e \u003cp\u003eCamera contrast is reported as camera present versus no camera, separated by public and private delivery points. In public delivery points, the role-gap OR was 1.39, indicating that bystanders weighted camera presence more favourably than recipients; consistent with cameras signalling accountability in shared spaces. In private delivery points, the direction reversed: the role-gap OR fell to 0.74, indicating that bystanders penalised cameras more than recipients near private property. This cross-context reversal indicates that privacy-related cues depend jointly on role and place. The urban and rural subsamples preserved this public\u0026ndash;private pattern (urban: 1.52 vs 0.73; rural: 1.25 vs 0.75). Confidence intervals were wider in rural areas and the rural public-delivery-point estimate did not reach statistical significance.\u003c/p\u003e \u003cp\u003eFor institutional safeguards, the pooled results show a modest licensing role gap, with bystanders placing slightly more weight on licensing than recipients. Registration showed little evidence of a role gap, suggesting broadly similar weighting across roles. The urban and rural subsamples were consistent with this picture: the licensing gap was clearest in the urban model while the rural estimate was less precise. Registration remained near parity in both settings. Taken together, safeguards mattered under both roles, but they did not separate bystanders from recipients as strongly as noise and cameras did.\u003c/p\u003e \u003cp\u003eOverall, the role shift changed preferences most for cameras and noise, while safety attributes were more stable across roles. Camera preferences showed a cross-context reversal: bystanders favoured cameras at public delivery points but penalised them near private ones, while recipients were largely indifferent. For noise, bystanders reacted more strongly than recipients in routine settings, but the two roles converged in noise-sensitive places. Licensing and registration were valued under both roles, with only modest role differences. These patterns were broadly consistent across urban and rural subsamples, though rural estimates were less precise for some contrasts. Recipient-side willingness-to-pay estimates, which translate these preference weights into monetary equivalents, are reported in Supplementary Information S7.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we show that acceptance of drone delivery is not captured by a single public preference. Existing evidence on societal acceptance often reports population-level attitudes or customer-side trade-offs, and it repeatedly identifies noise, privacy and safety as central concerns [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. By asking the same respondents to evaluate matched drone delivery scenarios as recipients and as bystanders, we find that encounter role changes the relative weight placed on these concerns. Bystanders penalised noisy operations more in everyday settings, but this role gap vanished in noise-sensitive locations, where recipients converged on similar noise sensitivity. Camera preferences showed a striking cross-context reversal: bystanders favoured cameras at public delivery points but penalised them near private ones, while recipients were largely indifferent. In contrast, institutional safeguards were valued under both roles, suggesting that licensing and registration operate as broadly shared assurance cues. These role-dependent patterns were broadly consistent across urban and rural respondents.\u003c/p\u003e \u003cp\u003eThese role gaps are consistent with a straightforward logic about who benefits and who bears the costs of a flight. Research on risk perception shows that judgements about new technologies depend on factors such as perceived lack of control and uncertainty, which can increase sensitivity to potential harms when people feel exposed rather than protected [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Trust in the organisations that manage risk also shapes how people judge hazards, especially when individuals cannot directly monitor operations or intervene [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. In our study, the bystander role approximates this low-control position: bystanders cannot opt out of overflight, gain no service benefit, and have no say over routing or timing. The recipient role, by contrast, makes the delivery benefit immediate and personal, which may offset concerns about noise or surveillance.\u003c/p\u003e \u003cp\u003eMore broadly, work on social acceptance of infrastructure argues that acceptance depends not only on average attitudes but on how local impacts and broader benefits are distributed across affected groups [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. When the costs of a technology fall disproportionately on those who do not share in its benefits acceptance among the cost-bearing group may erode even if aggregate sentiment is positive [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. For drone delivery, this distributive logic implies that \"public acceptance\" should be decomposed into at least two linked perspectives within the same people. This calls for service designs and governance frameworks that reduce impacts on non-recipients while making accountability visible to everyone.\u003c/p\u003e \u003cp\u003eNoise emerged as a robust constraint in our experiment, but its salience depended on both encounter role and operating context. Noisier flights reduced acceptance more strongly under the bystander viewpoint than under the recipient viewpoint, which fits the idea that external costs carry more weight when the service benefit is absent. When the scenario shifted to noise-sensitive places such as homes and hospitals, the role gap vanished, suggesting that tolerance narrows when delivery is imagined close to private life or vulnerable settings. This context-dependent noise sensitivity aligns with evidence from environmental acoustics showing that annoyance responses to the same sound level depend strongly on the activity being disrupted and the expectation of quiet in a given location [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. It is also consistent with public health guidance and epidemiological reviews that treat environmental noise as a material health risk and support targeted mitigation where exposure is most consequential [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. More specifically, psychoacoustic studies of small multirotor drones show that their tonal and high-frequency features generate disproportionate annoyance relative to conventional transport sources at equivalent exposure levels [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e], which may amplify the context effect we observe. For operators and regulators, the practical implication is to manage noise through both engineering and operations: prioritising lower-noise propulsion designs, reducing hovering and low-altitude manoeuvres near dwellings, and routing away from noise-sensitive land uses. Recipient willingness-to-pay estimates (Supplementary Information S7) reinforce that quieter operations carry tangible monetary value.\u003c/p\u003e \u003cp\u003eCamera use produced the clearest evidence that privacy concerns are context-specific and role-dependent. Prior work shows that privacy preferences can change sharply across situations, even for the same person, because privacy judgements depend on what information is collected, by whom, and in which setting [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. This helps interpret our cross-context pattern. Camera use did not act as a uniform privacy penalty. Under the bystander viewpoint, cameras were more acceptable in public delivery points, but were penalised at private delivery points, consistent with stronger privacy expectations near homes and private property. Under the recipient viewpoint, camera effects were comparatively small. Compared with studies that mainly measure customer-side privacy trade-offs for drone delivery, our results suggest that a single privacy effect is not stable across roles or settings. A practical implication is that camera policies should be framed as part of the service context, not as a fixed feature. Operators may reduce conflict by limiting data collection and retention by default (following GDPR), using clear purpose limitation, and adopting context-dependent controls such as geofenced camera restrictions [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn contrast, licensing and registration were valued under both roles. Recipient side evidence show that these safety measures carry the largest utility and monetary values (Supplementary information S7) among the attributes we tested. This pattern suggests that institutional safeguards function as broad assurance cues that support perceived accountability and the governance of risk, especially where individuals cannot directly monitor the system [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. The importance of identifiable and enforceable oversight is consistent with legitimacy research, which argues that rule following is closely tied to whether authorities and rules are seen as credible and justified. It also aligns with the direction of aviation governance, where identification tools are used to make operators traceable. For example, Remote ID in the United States and Europe is designed to provide identification and location information during flight and to help authorities respond to unsafe or unauthorised operations [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. While our study focuses on licensing and registration rather than Remote ID, the shared function of these attributes is to make responsibility clear when things go wrong. For implementation, the key is visibility. Operators and regulators can make licensing and registration easy to verify at the point of encounter through markings, simple public-facing identifiers, and app-based notifications. More generally, evidence from regulated sectors suggests that well-framed transparency can support citizen trust, although effects are modest and context-dependent [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. This reinforces the value of pairing formal safeguards with clear public communication that signals active oversight.\u003c/p\u003e \u003cp\u003eWe also examined whether the role-dependent patterns hold across settlement types, because everyday exposure and expectations about flight frequency can differ across urban, suburban and rural environments. Across all domains, the qualitative role-gap patterns replicated: noise and camera cues showed the strongest role and context dependence, while licensing and registration remained valued under both encounter roles. Beyond overall replication, one noteworthy area-level difference emerged for cameras. In public delivery points, the urban role-gap OR was above unity, while the rural estimate was smaller and its confidence interval included unity. This suggests that the bystander camera premium in public spaces may be more salient in denser urban environments where shared-space encounters are frequent, whereas rural respondents show more heterogeneous views. More broadly, several rural estimates had wider confidence intervals, consistent with prior research showing that perceived risks such as noise and privacy are central drivers of attitudes towards drone delivery, and that these concerns can be shaped by context and experience rather than being fixed traits [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eLimitations should be considered when interpreting the magnitudes of our estimates. First, the data come from an online survey in Poland (February to March 2022) with quota targets, which supports internal comparison but does not remove all risks of selection bias that can arise in web surveys [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. Second and similarly, as shown in Supplementary Information S1, our sample skews younger than the national population. This may bias responses toward individuals who are more technologically receptive, potentially underrepresenting the perspectives of older and less technology-confident groups. Third, choice tasks are stated preferences. They allow controlled trade-offs, but respondents may still behave differently when real services, prices and repeated exposure are involved [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. Fourth, encounter role is cued through framing. This is central to our design, but it also means that role differences can be influenced by how scenarios are described, and future field work could test whether similar gaps appear when people experience flights in real operating environments. Finally, while a mixed logit robustness check (Supplementary Information S6, supplementary table s7) confirms that the key role-dependent patterns are preserved after accounting for unobserved preference heterogeneity, our models do not explore systematic sources of this heterogeneity. For instance, whether the role gap varies by age, gender, income, or prior familiarity with drones. Future work could use latent class or hybrid choice specifications to identify whether particular population segments show larger or smaller role effects. Separately, the experimental design could be extended to disentangle specific camera functions (e.g., navigation vs recording), data retention policies, and flight frequency, which would help test how stable the patterns are as drone services move toward mass deployment.\u003c/p\u003e \u003cp\u003eDespite these caveats, the results point to a clear and actionable conclusion. Social acceptance of drone delivery is not a single attitude. It depends on where the flight occurs and on whether people experience the operation as recipients or as bystanders. This matters because drone delivery will only scale if services earn acceptance not just among customers but also among the wider communities in whose spaces flights take place [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Our findings imply that operators and regulators should treat noise and camera policies as context-sensitive design choices, while using licensing and registration as visible assurance tools that apply across roles. Designing for both encounter roles can strengthen accountability and improve the prospects for drone delivery that is compatible with diverse public expectations.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eSurvey design\u003c/p\u003e\n\u003cp\u003eThe questionnaire was designed to elicit preferences for drone delivery operations, with a focus on operational externalities and governance cues that are salient to public acceptance. The instrument combined sociodemographic and attitudinal questions with two stated preference discrete-choice experiments. Survey development followed an iterative process. First, two rounds of consultations with local authorities and other stakeholders were used to identify plausible drone delivery use cases and the operational issues most likely to shape acceptance. These inputs were translated into an initial stated preference experiment presented in a tabular format. Early testing indicated that respondents struggled to visualise the operating context from tables alone. To improve comprehension, we moved to a visual choice format. The visual stimuli were produced by graphic designers based on the attribute combinations specified in the experimental design. A second round of testing suggested that the visual format reduced respondent burden and improved clarity. Items that remained unclear were modified or removed before final deployment. The final survey was administered in Polish. Consent materials were also presented in Polish before the questionnaire began.\u003c/p\u003e\n\u003cp\u003eThe survey included two groups of stated preference discrete-choice experiments administered to the same individuals. Each choice task presented two unlabelled drone alternatives (Drone A and Drone B) and an opt-out option (\u0026ldquo;Neither\u0026rdquo;). Figures \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e include examples of a choice task presented to respondents, while the upper level shows the bystander tasks and the lower level presents the recipient tasks. The recipient experiment included service attributes that are only meaningful under the recipient framing (delivery time and delivery cost), while the bystander experiment included flight topography. In the primary pooled model, attributes common to both roles enter with recipient-baseline and bystander-shift parameters. Attributes unique to one role are incorporated through role-gated terms that restrict each to the relevant task block (with values set to zero outside the applicable role). This specification includes all attributes shown to respondents while preserving direct within-person comparability for the shared attributes. For recipient-side willingness-to-pay estimation (Supplementary Information S7, Supplementary Table S8, S9 \u0026amp; Fig. S3), we estimate a separate recipient-only model that includes delivery cost, scaled in 5 PLN units.\u003c/p\u003e\n\u003cp\u003eIn addition to the discrete-choice tasks, the questionnaire collected background information to describe the sample and to support subgroup analyses. Respondents reported sociodemographic characteristics (age, gender, household composition, employment status, dwelling type, and settlement type). Settlement type was used to define urban (n\u0026thinsp;=\u0026thinsp;599) and rural (n\u0026thinsp;=\u0026thinsp;402) subgroups for robustness checks. The questionnaire also collected measures of everyday delivery practices (delivery frequency, typical delivery locations, and approximate monthly spending on delivery services); these variables are not used in the choice models but provide descriptive context reported in Supplementary Information S1. Finally, we measured familiarity with drones (whether respondents had previously seen a drone) and attitudes toward drone delivery. The recipient block always preceded the bystander block, so encounter role is partially confounded with block order. The within-person design still enables direct comparison of role-framed valuations within respondents, but learning, anchoring, or fatigue may also contribute to the observed recipient\u0026ndash;bystander differences. We therefore interpret the estimates as role-framed within-person contrasts and note that future experiments should randomise block order to isolate pure role effects.\u003c/p\u003e\n\u003cp\u003eData collection\u003c/p\u003e\n\u003cp\u003eData were collected in Upper Silesian Region (GZM) in Poland in February\u0026ndash;March 2022 using an online questionnaire. A professional survey company was responsible for recruiting participants based on sample requirements (age, gender, and type of living area) to approximate census benchmarks for the study area. The survey remained in the field for approximately two weeks and the mean completion time was about 6 minutes, which is consistent with the task structure (13 choice tasks plus short background sections) and the visual format that simplified scenario comprehension. The analytical sample comprises 1,001 respondents who provided usable responses for the discrete-choice tasks. The recruitment contract specified delivery of completed questionnaires; accordingly, the dataset contains no incomplete choice tasks. The study received ethics approval from University College London (UCL) (Project Ethics Identification: 20751/001). All respondents provided informed consent prior to participation, and were compensated through the panel provider\u0026rsquo;s standard incentive scheme. During fieldwork, data were stored on an EU-based cloud server hosted by a project partner; after survey completion, the data were transferred to UCL and stored in the project\u0026rsquo;s shared repository for analysis. The dataset contains no identifiable personal information.\u003c/p\u003e\n\u003cp\u003eData preparation and analysis\u003c/p\u003e\n\u003cp\u003eThe bystander and recipient choice files were linked by respondent identifier and combined into a single pooled dataset. The choice outcome was coded as a three-category variable (Drone A, Drone B, or Neither). Binary attributes were coded as one-sided indicators with clear baselines: operator licensing (1\u0026thinsp;=\u0026thinsp;not licensed, 0\u0026thinsp;=\u0026thinsp;licensed), drone registration (1\u0026thinsp;=\u0026thinsp;not registered, 0\u0026thinsp;=\u0026thinsp;registered), and camera (1\u0026thinsp;=\u0026thinsp;no camera, 0\u0026thinsp;=\u0026thinsp;camera present). Noise was treated as an ordered five-level variable oriented so that a one-unit increase corresponds to one step noisier. Built context was recoded to a binary indicator (1\u0026thinsp;=\u0026thinsp;noise-sensitive: home or hospital; 0\u0026thinsp;=\u0026thinsp;other), and delivery point to a binary indicator (1\u0026thinsp;=\u0026thinsp;private: doorstep, terrace, window; 0\u0026thinsp;=\u0026thinsp;public/communal). Parcel type was recoded into three categories (medical, food, other as baseline). For recipient tasks, delivery cost was entered in 5 zł units. Role-gated variables (cost, time for recipients; topography dummies for bystanders) were set to zero outside the applicable role block to avoid missing-value propagation.\u003c/p\u003e\n\u003cp\u003eOur primary goal was to estimate the within-person change in attribute sensitivity between encounter roles. We analysed the pooled choice data using a panel multinomial logit model with three alternatives per task. Because each respondent completed repeated tasks across both role blocks, we treated the data as panel data and used respondent-clustered (panel-robust) standard errors.\u003c/p\u003e\n\u003cp\u003eTo quantify how encounter role reweights the same attributes, we used a role indicator r_nt coded as 0 for recipient tasks and 1 for bystander tasks. For shared attributes, coefficients under the recipient role serve as the baseline; bystander coefficients are expressed as the baseline plus an additive bystander shift. Attributes unique to one role enter through role-gated terms: delivery cost and time are multiplied by (1 \u0026ndash; r\u003csub\u003ent\u003c/sub\u003e) so they contribute only in recipient tasks, while topography dummies are multiplied by r_nt so they contribute only in bystander tasks. This parameterisation supports direct within-sample tests of role differences for each attribute and interaction term. Specifically, let respondent n\u0026thinsp;=\u0026thinsp;1,\u0026hellip;, N evaluate task t\u0026thinsp;=\u0026thinsp;1,\u0026hellip;,T\u003csub\u003ent\u003c/sub\u003e over three alternatives j\u0026isin;{A,B,N} (where N denotes \u0026ldquo;Neither\u0026rdquo;). Let r\u003csub\u003ent\u003c/sub\u003e\u0026isin;{0,1} be the role indicator (0\u0026thinsp;=\u0026thinsp;recipient viewpoint, 1\u0026thinsp;=\u0026thinsp;bystander viewpoint). This parameterisation means coefficients under the bystander viewpoint are the sum of the recipient baseline and the bystander shift. Utilities are alternative-specific and linear in attributes. We write utility as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{U}_{ntj}=\\:{V}_{ntj}+\\:{\\epsilon\\:}_{ntj}\\)\u003c/span\u003e\u003c/span\u003e, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\epsilon\\:}_{ntj}\\:\\)\u003c/span\u003e\u003c/span\u003erepresenting type I extreme value. Under the i.i.d. Type I extreme value assumption, the probability that respondent n chooses alternative j in task t is:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eWe estimate a pooled multinomial logit with role-specific slopes and context interactions. The role indicator r\u003csub\u003en\u003c/sub\u003e equals 1 when the same participant evaluates scenarios as a bystander and 0 when evaluating as a recipient (two viewpoints on one sample). The vector of drone attributes for alternative \u003cem\u003ej\u003c/em\u003e in task \u003cem\u003et\u003c/em\u003e comprises:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNoise\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026ndash; Ordered noise score, coded so that a one-unit increase is one step noisier (louder);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSense\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026ndash; built-context indicator (1\u0026thinsp;=\u0026thinsp;noise-sensitive: home or hospital; 0\u0026thinsp;=\u0026thinsp;other);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eLic\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e\u0026isin;{0,1} \u0026ndash; licensing coding (1\u0026thinsp;=\u0026thinsp;not licensed, 0\u0026thinsp;=\u0026thinsp;licensed);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eReg\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026isin;{0,1} \u0026ndash; registration coding (1\u0026thinsp;=\u0026thinsp;not registered, 0\u0026thinsp;=\u0026thinsp;registered);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCam\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026isin;{0,1} \u0026ndash; camera coding used in estimation (1\u0026thinsp;=\u0026thinsp;no camera, 0\u0026thinsp;=\u0026thinsp;camera present);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePriv\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026isin;{0,1} \u0026ndash; delivery point indicator (1\u0026thinsp;=\u0026thinsp;private: doorstep, terrace, window; 0\u0026thinsp;=\u0026thinsp;public/communal);\u003c/p\u003e\n\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u0026nbsp;\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:\\)\u003c/span\u003e\u0026nbsp;\u003c/span\u003e \u003csub\u003e\u003cem\u003eN,rec\u003c/em\u003e\u003c/sub\u003e \u0026ndash; opt-out intercept under the recipient viewpoint;\u003c/p\u003e\n\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u0026nbsp;\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:\\)\u003c/span\u003e\u0026nbsp;\u003c/span\u003e \u003csub\u003e\u003cem\u003eN,by\u003c/em\u003e\u003c/sub\u003e \u003cem\u003e\u0026ndash;\u003c/em\u003e additive opt-out shift under the bystander viewpoint;\u003c/p\u003e\n\u003cp\u003eIn addition, the following role-gated attributes enter utility only for the relevant role:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCost\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026mdash; delivery cost in 5 zł units (recipient tasks only; set to zero for bystander tasks);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTime\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e\u0026mdash; delivery time (recipient tasks only; set to zero for bystander tasks);\u003c/p\u003e\n\u003cp\u003eVillage\u003csub\u003e\u003cem\u003entj\u003c/em\u003e\u003c/sub\u003e \u0026isin; {0,1} \u0026mdash; topography indicator for over-village flight (bystander tasks only; set to zero for recipient tasks; baseline\u0026thinsp;=\u0026thinsp;over-city);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRural\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003entj\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u0026isin; {0,1} \u0026mdash; topography indicator for rural-area flight (bystander tasks only; set to zero for recipient tasks; baseline\u0026thinsp;=\u0026thinsp;over-city);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eParcel\u003c/em\u003e_med\u003csub\u003e\u003cem\u003entj\u003c/em\u003e\u003c/sub\u003e \u0026isin; {0,1} \u0026mdash; parcel content indicator for medical deliveries (both roles; baseline\u0026thinsp;=\u0026thinsp;other);\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eParcel\u003c/em\u003e_food\u003csub\u003e\u003cem\u003entj\u003c/em\u003e\u003c/sub\u003e \u0026isin; {0,1} \u0026mdash; parcel content indicator for food deliveries (both roles; baseline\u0026thinsp;=\u0026thinsp;other).\u003c/p\u003e\n\u003cp\u003eNoise is entered linearly as an ordered 1\u0026ndash;5 scale, which imposes a constant marginal effect per step increase in noisiness; we use this parsimonious specification because the stimulus was designed as a uniform step scale. For camera, operator licensing, and drone registration, the model uses \u0026quot;undesirable-state\u0026quot; indicator variables (Cam\u0026thinsp;=\u0026thinsp;1 when no camera is present, Lic\u0026thinsp;=\u0026thinsp;1 when not licensed, Reg\u0026thinsp;=\u0026thinsp;1 when not registered). In Results, we report contrasts oriented toward the desirable state (e.g., \u0026quot;camera present vs no camera,\u0026quot; \u0026quot;licensed vs not licensed\u0026quot;). These correspond to a one-unit decrease in the respective indicator (i.e., moving from the undesirable to the desirable state), so the reported odds ratios equal exp(\u0026minus;\u0026beta;) for each attribute.\u003c/p\u003e\n\u003cp\u003eBecause Drone A and Drone B are unlabelled (generic) alternatives, we normalise their alternative-specific constants to zero, only the opt-out (\u0026quot;Neither\u0026quot;) receives an estimated constant (with a recipient baseline and a bystander shift). This is standard for unlabelled designs where the alternatives carry no inherent identity beyond their attributes [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. We verified that introducing separate left/right position constants did not materially change the results. The urban/rural estimation follows the same model structure, estimated separately within urban (n\u0026thinsp;=\u0026thinsp;599) and rural (n\u0026thinsp;=\u0026thinsp;402) subsamples. Because separate estimation does not constrain the error scale to be equal across subsamples, we interpret cross-area comparisons as pattern-replication checks rather than formal magnitude tests.\u003c/p\u003e\n\u003cp\u003eTo assess whether the pooled MNL results depend on the assumption of homogeneous preferences, we re-estimated the pooled role specification as a mixed logit with normally distributed random coefficients on all four core shared attributes (noise, camera, licensing, and registration). Bystander shifts and context interactions were kept as fixed parameters. The model was estimated by maximum simulated likelihood with 500 Halton draws and panel likelihood across repeated tasks. We report the mixed logit as a robustness check in Supplementary Information S6. The MNL is retained as the primary specification because it provides a more transparent baseline for interpreting role and context interactions, and the mixed logit confirms that the substantive role-gap patterns are preserved after accounting for unobserved heterogeneity.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eData availability \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are not publicly available due to ethics restrictions, but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003eCode availability\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe R scripts used for computing the results in this manuscript can be found in below github redispository (https://github.com/Basilcao/Drone-manuscript).\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe research reported in this paper has been supported by European Union’s Horizon 2020 research and innovation programme under Grant Agreement No 815269, project HARMONY. The contribution of Luciano Pana Tronca is acknowledged for working with GZM for the workshops that fed with insights the questionnaire design.\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMK perceived the idea of this survey. MK, FS and YZ performed study design, data curation and preparation; YC, MK, FS and YZ analyzed and interpreted the data. YC collaborate with the graphic designers for the visualisation of the SPs, prepared all the statistical figures and was a major contributor in writing the manuscript. MK, FS and YZ supervised, reviewed and revised. All author wrote, read, and approved the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe author declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eJ.K. Stolaroff, C. Samaras, E.R. O\u0026rsquo;Neill, A. Lubers, A.S. Mitchell, D. Ceperley, Energy use and life cycle greenhouse gas emissions of drones for commercial package delivery, Nature communications 9(1) (2018) 409.\u003c/li\u003e\n \u003cli\u003eA. Raghunatha, P. Hilletofth, P. Thollander, Exploring the dynamics of public acceptance factors and drone transport planning towards sustainable transitions, European Transport Research Review 17(1) (2025) 1-14.\u003c/li\u003e\n \u003cli\u003eV. Garg, S. Niranjan, V. Prybutok, T. Pohlen, D. Gligor, Drones in last-mile delivery: A systematic review on Efficiency, Accessibility, and Sustainability, Transportation Research Part D: Transport and Environment 123 (2023) 103831.\u003c/li\u003e\n \u003cli\u003eA. Zenz, J. Powles, Resisting technological inevitability: Google Wing\u0026rsquo;s delivery drones and the fight for our skies, Philosophical Transactions A 382(2285) (2024) 20240107.\u003c/li\u003e\n \u003cli\u003eS. Zhang, P. Anciaes, E. Chaniotakis, The future of urban deliveries? Perceptions and intentions about delivery robots and drones in Europe, Springer2025.\u003c/li\u003e\n \u003cli\u003eR. W\u0026uuml;stenhagen, M. Wolsink, M.J. B\u0026uuml;rer, Social acceptance of renewable energy innovation: An introduction to the concept, Energy policy 35(5) (2007) 2683-2691.\u003c/li\u003e\n \u003cli\u003eC. Al Haddad, E. Chaniotakis, A. Straubinger, K. Pl\u0026ouml;tner, C. Antoniou, Factors affecting the adoption and use of urban air mobility, Transportation research part A: policy and practice 132 (2020) 696-712.\u003c/li\u003e\n \u003cli\u003eQ. Long, J. Ma, F. Jiang, C.J. Webster, Demand analysis in urban air mobility: A literature review, Journal of Air Transport Management 112 (2023) 102436.\u003c/li\u003e\n \u003cli\u003eN. EASA, Study on the societal acceptance of urban air mobility in Europe, European Union Aviation Safety Agency and McKinsey \u0026amp; Company (2021).\u003c/li\u003e\n \u003cli\u003eS. Hess, K. Train, Correlation and scale in mixed logit models, Journal of choice modelling 23 (2017) 1-8.\u003c/li\u003e\n \u003cli\u003eE. Lancsar, J. Louviere, Deleting \u0026lsquo;irrational\u0026rsquo;responses from discrete choice experiments: a case of investigating or imposing preferences?, Health economics 15(8) (2006) 797-811.\u003c/li\u003e\n \u003cli\u003eB.I. Scott, K.I. Andritsos, A drone strategy 2.0 for a smart and sustainable unmanned aircraft eco-system in Europe, Air and Space Law 48(3) (2023).\u003c/li\u003e\n \u003cli\u003eW. Yoo, E. Yu, J. Jung, Drone delivery: Factors affecting the public\u0026rsquo;s attitude and intention to adopt, Telematics and Informatics 35(6) (2018) 1687-1700.\u003c/li\u003e\n \u003cli\u003eN. Wang, N. Mutzner, K. Blanchet, Societal acceptance of urban drones: A scoping literature review, Technology in Society 75 (2023) 102377.\u003c/li\u003e\n \u003cli\u003eR. Khan, S. Tausif, A. Javed Malik, Consumer acceptance of delivery drones in urban areas, International Journal of Consumer Studies 43(1) (2019) 87-101.\u003c/li\u003e\n \u003cli\u003eR. Kellermann, T. Biehle, H. Mostofi, Modelling public attitude towards drone delivery in Germany, European Transport Research Review 15(1) (2023) 38.\u003c/li\u003e\n \u003cli\u003eA.W. Christian, R. Cabell, Initial investigation into the psychoacoustic properties of small unmanned aerial system noise, 23rd AIAA/CEAS aeroacoustics conference, 2017, p. 4051.\u003c/li\u003e\n \u003cli\u003eS.H. Kim, Choice model based analysis of consumer preference for drone delivery service, Journal of Air Transport Management 84 (2020) 101785.\u003c/li\u003e\n \u003cli\u003eWHO, Environmental noise guidelines for European region, World Health Organization, 2019.\u003c/li\u003e\n \u003cli\u003eP.J. Lee, C. Hopkins, R. Penedo, Attitudes to noise inside dwellings in three megacities: Seoul, London, and S\u0026atilde;o Paulo, International Journal of Environmental Research and Public Health 17(16) (2020) 6005.\u003c/li\u003e\n \u003cli\u003eS. Leon, C. Chen, A. Ratcliffe, Consumers\u0026rsquo; perceptions of last mile drone delivery, International Journal of Logistics Research and Applications 26(3) (2023) 345-364.\u003c/li\u003e\n \u003cli\u003eR. Altawy, A.M. Youssef, Security, privacy, and safety aspects of civilian drones: A survey, ACM Transactions on Cyber-Physical Systems 1(2) (2016) 1-25.\u003c/li\u003e\n \u003cli\u003eY. Wang, H. Xia, Y. Huang, Examining American and Chinese internet users\u0026apos; contextual privacy preferences of behavioral advertising, Proceedings of the 19th ACM conference on computer-supported cooperative work \u0026amp; social computing, 2016, pp. 539-552.\u003c/li\u003e\n \u003cli\u003eM. Siegrist, Trust and risk perception: A critical review of the literature, Risk analysis 41(3) (2021) 480-490.\u003c/li\u003e\n \u003cli\u003eS.A. Applin, Deliveries by drone: Obstacles and sociability, The Future of Drone Use: Opportunities and Threats from Ethical and Legal Perspectives, Springer2016, pp. 71-91.\u003c/li\u003e\n \u003cli\u003eA. AlHogail, Improving IoT technology adoption through improving consumer trust, Technologies 6(3) (2018) 64.\u003c/li\u003e\n \u003cli\u003eE. \u0026Ouml;z, E. Heikkil\u0026auml;, R. Tiusanen, Development of an Organisational Certification Process for Specific Category Drone Operations, Drones 6(10) (2022) 278.\u003c/li\u003e\n \u003cli\u003eY. Zhang, M. Kamargianni, A review on the factors influencing the adoption of new mobility technologies and services: autonomous vehicle, drone, micromobility and mobility as a service, Transport reviews 43(3) (2023) 407-429.\u003c/li\u003e\n \u003cli\u003eS.N. Lingam, R. Verstegen, S.M. Petermeijer, M. Martens, Human interactions with delivery drones in public spaces: design recommendations from recipient and bystander perspectives, Frontiers in Robotics and AI 12 (2025) 1580289.\u003c/li\u003e\n \u003cli\u003eA. Straubinger, U. Kluge, M. Fu, C. Al Haddad, K.O. Ploetner, C. Antoniou, Identifying demand and acceptance drivers for user friendly urban air mobility introduction, Towards User-Centric Transport in Europe 2: Enablers of Inclusive, Seamless and Sustainable Mobility, Springer2020, pp. 117-134.\u003c/li\u003e\n \u003cli\u003eY.-J. Tu, S. Piramuthu, Security and privacy risks in drone-based last mile delivery, European Journal of Information Systems 33(5) (2024) 617-630.\u003c/li\u003e\n \u003cli\u003eQ. Tan, H. Bian, J. Guo, P. Zhou, H.K. Lo, S. Zhong, X. Zhang, Virtual flight simulation of delivery drone noise in the urban residential community, Transportation Research Part D: Transport and Environment 118 (2023) 103686.\u003c/li\u003e\n \u003cli\u003eL.K.L. Tan, B.C. Lim, G. Park, K.H. Low, V.C.S. Yeo, Public acceptance of drone applications in a highly urbanized environment, Technology in Society 64 (2021) 101462.\u003c/li\u003e\n \u003cli\u003eL.Y. Koh, J.Y. Lee, X. Wang, K.F. Yuen, Urban drone adoption: Addressing technological, privacy and task\u0026ndash;technology fit concerns, Technology in Society 72 (2023) 102203.\u003c/li\u003e\n \u003cli\u003eJ.J. Louviere, D.A. Hensher, J.D. Swait, Stated choice methods: analysis and applications, Cambridge university press2000.\u003c/li\u003e\n \u003cli\u003eM. Kamargianni, S. Dubey, A. Polydoropoulou, C. Bhat, Investigating the subjective and objective factors influencing teenagers\u0026rsquo; school travel mode choice\u0026ndash;An integrated choice and latent variable model, Transportation Research Part A: Policy and Practice 78 (2015) 473-488.\u003c/li\u003e\n \u003cli\u003eS. Hess, D. Palma, Apollo: A flexible, powerful and customisable freeware package for choice model estimation and application, Journal of choice modelling 32 (2019) 100170.\u003c/li\u003e\n \u003cli\u003eT. Islam, J.J. Louviere, P.F. Burke, Modeling the effects of including/excluding attributes in choice experiments on systematic and random components, International Journal of Research in Marketing 24(4) (2007) 289-300.\u003c/li\u003e\n \u003cli\u003eD.G. Fiebig, M.P. Keane, J. Louviere, N. Wasi, The generalized multinomial logit model: accounting for scale and coefficient heterogeneity, Marketing science 29(3) (2010) 393-421.\u003c/li\u003e\n \u003cli\u003eJ. Swait, J. Louviere, The role of the scale parameter in the estimation and comparison of multinomial logit models, Journal of marketing research 30(3) (1993) 305-314.\u003c/li\u003e\n \u003cli\u003eA. Berke, G. Ding, C. Chin, K. Gopalakrishnan, K. Larson, H. Balakrishnan, M.Z. Li, Drone delivery and the value of customer privacy: A discrete choice experiment with US consumers, Transportation Research Part C: Emerging Technologies 157 (2023) 104391.\u003c/li\u003e\n \u003cli\u003eP. Slovic, Perception of bisk, Emerging Technologies, Routledge2020, pp. 141-146.\u003c/li\u003e\n \u003cli\u003eR. Guski, D. Schreckenberg, R. Schuemer, WHO environmental noise guidelines for the European region: A systematic review on environmental noise and annoyance, International journal of environmental research and public health 14(12) (2017) 1539.\u003c/li\u003e\n \u003cli\u003eT. M\u0026uuml;nzel, F.P. Schmidt, S. Steven, J. Herzog, A. Daiber, M. S\u0026oslash;rensen, Environmental noise and the cardiovascular system, Journal of the American College of Cardiology 71(6) (2018) 688-697.\u003c/li\u003e\n \u003cli\u003eC. Kawai, J. J\u0026auml;ggi, F. Georgiou, J. Meister, R. Pieren, B. Sch\u0026auml;ffer, Short-term noise annoyance towards drones and other transportation noise sources: A laboratory study, The Journal of the Acoustical Society of America 156(4) (2024) 2578-2595.\u003c/li\u003e\n \u003cli\u003eN.S. Zawodny, A. Christian, R. Cabell, A summary of NASA research exploring the acoustics of small unmanned aerial systems, 2018 AHS Technical Meeting on Aeromechanics Design for Transformative Vertical Flight, 2018.\u003c/li\u003e\n \u003cli\u003eA. Acquisti, L. Brandimarte, G. Loewenstein, Privacy and human behavior in the age of information, Science 347(6221) (2015) 509-514.\u003c/li\u003e\n \u003cli\u003eH. Nissenbaum, Privacy as contextual integrity, Wash. L. Rev. 79 (2004) 119.\u003c/li\u003e\n \u003cli\u003eE. Commission, Commission Implementing Regulation (EU) 2019/947 of 24 May 2019 on the rules and procedures for the operation of unmanned aircraft, Official Journal of the European Union L 152, EU, 2019.\u003c/li\u003e\n \u003cli\u003eU.S.G.A. Office, Drones: Actions needed to better support remote identification in the national airspace (GAO-24-106158), in: GAO (Ed.) Washington, DC, 2024.\u003c/li\u003e\n \u003cli\u003eS. Grimmelikhuijsen, F. de Vries, R. Bouwman, Regulators as guardians of trust? The contingent and modest positive effect of targeted transparency on citizen trust in regulated sectors, Journal of Public Administration Research and Theory 34(1) (2024) 136-149.\u003c/li\u003e\n \u003cli\u003eJ. Bethlehem, Selection bias in web surveys, International statistical review 78(2) (2010) 161-188.\u003c/li\u003e\n \u003cli\u003eR.T. Carson, T. Groves, Incentive and informational properties of preference questions, Environmental and resource economics 37(1) (2007) 181-210.\u003c/li\u003e\n \u003cli\u003eJ.J. Murphy, P.G. Allen, T.H. Stevens, D. Weatherhead, A meta-analysis of hypothetical bias in stated preference valuation, Environmental and Resource Economics 30(3) (2005) 313-325.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-sustainable-mobility-and-transport","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [npj Sustainable Mobility and Transport](https://www.nature.com/npjsustainmobil)","snPcode":"44333","submissionUrl":"https://submission.springernature.com/new-submission/44333/3","title":"npj Sustainable Mobility and Transport","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9412457/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9412457/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDrone delivery promises faster, cleaner logistics but raises concerns about noise, privacy, and safety that differ depending on how people encounter the technology \u0026mdash; whether as recipients who benefit directly or as bystanders who bear externalities. Current research rarely distinguishes between these perspectives, overlooking how preferences may depend on the context of that encounter. Using a discrete-choice experiment in Poland (N\u0026thinsp;=\u0026thinsp;1,001) in which the same individuals evaluated matched scenarios from both perspectives, we estimate a pooled multinomial logit model with role-specific interactions. Results show role-dependent trade-offs. Bystanders penalise noisier operations more than recipients in routine settings, but this gap vanishes in noise-sensitive locations. Meanwhile, camera preferences are strongly context-dependent: bystanders favour cameras at public delivery points but resist them near private ones, while recipients show weak camera sensitivity overall. Both roles value operator licensing and drone registration, with bystanders placing modestly greater emphasis on compliance. These findings offer guidance for tailoring drone service design and communication to different publics.\u003c/p\u003e","manuscriptTitle":"Bystander and recipient roles reweight noise, privacy, and safety in drone deliveries","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-05 11:03:50","doi":"10.21203/rs.3.rs-9412457/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-01T13:48:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"272363629433926075403319273823486757456","date":"2026-05-01T13:43:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"52119610996369526228318300082432819708","date":"2026-04-30T07:21:40+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-24T09:37:30+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-24T09:32:48+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-24T02:57:37+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Sustainable Mobility and Transport","date":"2026-04-14T08:06:01+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-sustainable-mobility-and-transport","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [npj Sustainable Mobility and Transport](https://www.nature.com/npjsustainmobil)","snPcode":"44333","submissionUrl":"https://submission.springernature.com/new-submission/44333/3","title":"npj Sustainable Mobility and Transport","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"bfc0d096-6aae-4fc2-b160-4615ca4031db","owner":[],"postedDate":"May 5th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-01T13:48:28+00:00","index":53,"fulltext":""},{"type":"reviewerAgreed","content":"272363629433926075403319273823486757456","date":"2026-05-01T13:43:42+00:00","index":52,"fulltext":""},{"type":"reviewerAgreed","content":"52119610996369526228318300082432819708","date":"2026-04-30T07:21:40+00:00","index":49,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":67513115,"name":"Physical sciences/Engineering"},{"id":67513116,"name":"Biological sciences/Psychology"},{"id":67513117,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-05-05T11:03:50+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-05 11:03:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9412457","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9412457","identity":"rs-9412457","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}