Pith. sign in

REVIEW 3 major objections 5 minor 34 references

Towards Developing Socially Compliant Automated Vehicles: Advances, Expert Insights, and A Conceptual Framework

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper proposes a five-module framework for socially compliant automated vehicles and reports survey evidence from 90 professionals endorsing it.

desk verdict Useful organizing framework for SCAV research; the survey 'validation' is an endorsement poll and should be read as stakeholder feedback, not evidence. read the letter →

arxiv 2501.06089 v3 pith:ADQV56PC submitted 2025-01-10 cs.RO cs.AIcs.LGcs.MAcs.SYeess.SY

Yongqi Dong , Bart van Arem , Haneen Farah This is my paper · ORCID
classification cs.ROcs.AIcs.LGcs.MAcs.SYeess.SY
keywords Automatedvehicles(AVs)SociallycompliantdrivingMixedtrafficConceptualframeworkScopingreviewExpertsurveyHuman-drivenBidirectionalbehavioraladaptation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to establish that automated vehicles entering mixed traffic need a dedicated engineering agenda for social compliance, and that the authors' five-module conceptual framework is the right organising structure for that agenda. The paper's evidence chain runs from a scoping review of 68 studies, through interviews with ten experts, to a survey of 90 professionals who rate the framework's nine technical capabilities and rank their priorities. The authors use the survey results to claim that all framework elements are seen as significant, with anticipation of other road users' actions rated most urgent for the near term and bidirectional behavioural adaptation plus spatial-temporal memory rated most critical for the long term. If the framework is accepted, it would turn 'socially compliant driving' from a scattered research theme into a coordinated development roadmap for industry, academia, and policymakers.

What carries the argument

The carrying object is the five-module conceptual framework shown in the paper's Fig. 5. Each module is tied to a specific limitation identified in current AVs: excessive conservatism, inability to read implicit communications, poor adaptation to driving styles, limited scenario anticipation, and cultural inflexibility. The framework also includes standard sensing/perception and communication/eHMI modules, and the survey instrument translates the framework into nine rated technical capabilities, giving the conceptual structure an operational form that experts could judge. That translation from framework modules to survey items is the mechanism by which the paper converts expert opinion into evidence for the framework.

What would settle it

If a replication survey with balanced representation from North America, Japan, and India returned an average importance rating at or below the neutral midpoint for any of the nine capabilities, the paper's claim that all framework elements are significant would be contradicted. A sharper test would be a controlled driving experiment in which an AV implementing the full framework is compared with a conventional AV; if human drivers do not predict the framework AV's actions more accurately in merging and left-turn scenarios, the claim that these modules enable social compliance would fail.

Watch

Extended reading notes

Core claim

The central claim is that socially compliant automated driving is best understood and advanced through one integrated framework rather than through isolated algorithms. The proposed framework has five interdependent modules: a socially compliant decision-making module that embeds culture, norms, implicit cues, and driving styles; a safety constraint module that enforces hard safety boundaries on every planned action; an ego-versus-network trade-off module that balances the ego vehicle's benefits against network-level efficiency and societal outcomes; a bidirectional behavioural adaptation module through which AVs and human drivers adjust to each other over time; and a spatial-temporal memory module that stores short- and long-term interaction histories to refine future decisions. The paper treats the 90-respondent expert survey as validation of this structure: all nine rated technical capabilities score above the neutral midpoint of the rating scale, and the ranking exercises place anticipation capability first for medium-term development and bidirectional behavioural adaptation with spatial-temporal memory first for the long term. The contribution is therefore an integrated conceptual map and a prioritised research agenda, with the survey presented as evidence that the map matches expert expectations.

Load-bearing premise

The framework's generalisability rests on the assumption that the 90 survey respondents, drawn mostly from Europe and China with only one from the United States, represent the global range of driving cultures and expert priorities; if they do not, the framework's capability set and timing could be regionally biased.

Editorial extensions

If this is right

  • If the framework's priority ordering is right, near-term SCAV development should concentrate on anticipation capability, the ability to read other road users' intended actions, before investing heavily in other social features.
  • Long-term research and development should shift toward bidirectional behavioural adaptation and spatial-temporal memory, which experts ranked as the most critical capabilities for the 5-10 year horizon.
  • Safety should be enforced by a dedicated module that sits outside the socially compliant decision-making layer, meaning social behaviour is constrained but not determined by safety logic.
  • Social compliance should be treated as a dynamic trade-off between ego-vehicle benefits and network-level outcomes, so optimal individual behaviour is not assumed to be optimal for the road network.
  • The five methodological families identified in the scoping review, including imitation learning, reinforcement learning with utility models, game-theoretic and field-based models, trajectory prediction, and optimisation-based tuning, are complementary and should be combined in future SCAV systems.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the framework generalises, eHMI design and cultural calibration become first-class engineering requirements rather than optional extras; a testable extension would be matched on-road trials in two countries with different implicit driving norms to see whether the same module weights predict acceptance.
  • The survey's geographic imbalance means the priority ordering may shift with a more global sample; a direct test would re-run the ranking questions with balanced representation from North America, Japan, and India and compare rank orderings.
  • Because the framework is conceptual, its next test is computational instantiation; one concrete check is whether the spatial-temporal memory module improves long-horizon merging or left-turn decisions in simulation relative to a memory-free baseline.
  • The framework's bidirectional adaptation implies AV deployment is a co-evolution process: as human drivers learn to exploit or trust AVs, AV policies should update in response, which would make field deployment a continuous calibration loop rather than a one-time release.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper presents a scoping review of research on socially compliant automated vehicles (SCAVs), supplemented by informal expert interviews, and proposes a conceptual framework comprising five elements: socially compliant decision-making, a safety constraint module, ego-versus-network trade-off management, bidirectional behavioral adaptation, and spatial-temporal memory. The framework is then 'evaluated' with an online survey of 90 experts who rated and ranked the importance of nine technical capabilities that correspond directly to the framework's components. The paper claims that the survey results provide valuable validation and affirm the significance of the framework. The scoping review and the qualitative synthesis are systematic, and the authors share their data and code. However, the survey-based validation is non-discriminating: it measures endorsement of the framework's own components without control items, statistical inference, or a formal analysis of open-ended suggestions, so it cannot substantiate the strong validation claims made in the Abstract and Section 4.3.1.

Significance. If the framework is viewed as a synthesis of the literature and expert opinion, it offers a useful structuring of an emerging field and a plausible research agenda. The scoping review is methodical, follows a PRISMA-based process, and covers a broad range of methods, datasets, and scenarios; the replication data and code are a concrete strength. The proposed capabilities (e.g., bidirectional adaptation, spatial-temporal memory) are grounded in gaps identified in the review and the expert interviews. The paper's main weakness is the evidential weight placed on the survey: the ratings and rankings are consistent with the framework by construction, so they add little beyond the initial synthesis. The significance of the paper therefore rests on the review and framework rather than on the claimed empirical validation. With a more modest interpretation of the survey and explicit acknowledgment of its limits, the contribution is valuable.

major comments (3)
  1. [Section 4.3.1, Fig. 10] The survey asks respondents to rate exactly the nine capabilities that compose Fig. 5, and the average rating for every capability exceeds 4.8 on a 1–7 scale. The text states that this result 'supports and verifies the elements proposed in the conceptual framework.' This inference is not justified: because the items were derived from the framework itself, the high ratings show only that experts endorse the framework's components, not that the component set is correct, complete, or better than an alternative. No control items are included, no statistical tests are performed, and there is no comparison with a rival framework. Please either add a discriminant check (e.g., items that are intentionally less central, or a comparison with an alternative component list) or reframe the claim as 'expert endorsement' rather than 'validation.'
  2. [Section 4.3.1, Figs. 11–12; Section 4.3.3] The ranking tasks restrict respondents to preselected capabilities from the framework, so they cannot reveal an omitted high-priority capability. The open-ended question 'What else would you expect for the socially compliant automated vehicles?' is collected but only summarized qualitatively, with no coding scheme, inter-rater reliability, or a specified threshold for when a suggestion would count as a missing framework element. The central claim that the framework 'captures the key elements needed' (Section 3.2) is therefore not tested for completeness. Please perform a formal content analysis of the open-ended responses and report whether any suggestions fall outside the framework, or explicitly state that completeness remains an open question.
  3. [Section 5, 'Limitations' paragraph; Abstract] The authors acknowledge the geographic imbalance in the survey, but the Abstract and Section 4.3.1 still state that the survey provides 'valuable validation and insights, affirming the significance' of the framework. Beyond geography, the sample is self-selected and heavily weighted toward researchers (49 of 90 respondents), with only two OEM developers and seven policymakers. The survey also lacks any inferential statistics. These features severely constrain what the survey can establish, even beyond the acknowledged underrepresentation of the USA. Please moderate the validation claims throughout the manuscript to match what an exploratory, non-representative opinion survey can support, and explicitly list the sample's occupational skew and self-selection as limitations.
minor comments (5)
  1. [Table 3] The entry 'Bayesian ınference' contains a dotless 'ı'; correct this typo to 'inference.'
  2. [Section 2.2.1] The flow diagram indicates 1327 unique records after duplicate removal, but the text says 'Together with the manual examination of the titles, a total of 1327 valid unique records were included.' Please clarify whether title screening occurred before or after deduplication, so the reader can follow the PRISMA pipeline precisely.
  3. [Section 4.1 and Fig. 6 note] The explanation that the USA is grouped under 'Other' appears both in the main text and in the figure note. Move the full explanation to one place to reduce redundancy.
  4. [Supplementary attachments] Supplementary Attachment 1 (referenced in Section 2.2.1) and Supplementary Attachment 2 (referenced in Section 4) are both listed under the same shortened URL (https://lnkd.in/gpceU6gQ). Please verify that the link points to the correct documents or provide distinct URLs.
  5. [Introduction] The definition of socially compliant driving is clear, but consider adding a one-sentence definition in the Abstract or Nomenclature so that the term is immediately understandable to readers who skip the Introduction.

Circularity Check

2 steps flagged · score 6.0 of 10

The survey-based 'validation' is self-referential: the nine rated/ranked capabilities are the framework's own components, so high ratings and 'alignment' outcomes do not independently verify the framework.

  1. self definitional [Section 4.3.1, 'Rating and ranking of the identified key technical capabilities' (text around Fig. 10)]
    "Corresponding to the developed conceptual framework (Fig. 5), the respondents were asked to rate 9 key technical capabilities on a scale from 1 to 7... All 9 key technical capabilities were rated as significant, with average ratings exceeding 4.8, which supports and verifies the elements proposed in the conceptual framework (Fig. 5)."

    The validation instrument is constructed from the very framework it is used to validate: the nine rated items are the framework's own components ('Corresponding to the developed conceptual framework'). Treating high mean ratings of those same items as 'support[ing] and verify[ing] the elements proposed in the conceptual framework' is a self-consistency check, not an external test. No control or alternative capabilities are included, no threshold is specified for a missing element, and no rival framework is compared, so generic acquiescence or the social desirability of normatively plausible capabilities would produce the same result. The conclusion therefore reduces to the input list by construction.

  2. other [Section 4.3.1, long-term development ranking (text around Fig. 12)]
    "The results revealed that bidirectional behavioral adaptation ranked first, followed by spatial-temporal memory buffer integration, which is reasonable and aligns well with the proposed conceptual framework in Fig. 5."

    The long-term ranking question forces respondents to choose among four capabilities preselected from the proposed framework, and the resulting order is then reported as 'align[ing] well with the proposed conceptual framework'. Because the option set is the framework itself, any ranking of those options is consistent with the framework; the exercise cannot reveal an omitted high-priority capability or test the framework against an alternative. The 'alignment' outcome is therefore a restatement of the preselected input set rather than independent confirmation.

full rationale

The paper's scoping review and expert interviews provide independent content: the framework is proposed 'By incorporating insights from the scoping review and addressing the identified gaps and research expectations from both the literature review and the expert interviews' (Section 3.2), and the review follows a structured five-step PRISMA-style process. The circularity is localized to the validation step. In Section 4.3.1, the rating and ranking items are explicitly generated 'Corresponding to the developed conceptual framework (Fig. 5)', and high ratings plus forced rankings are then interpreted as supporting and verifying the framework. That makes the survey validation self-referential and unable to discriminate the proposed framework from plausible rivals or to establish completeness. The acknowledged geographic imbalance of the sample is a separate generalizability limitation, not a circularity. Because the independent review content remains substantive but the central 'validation' claim is partly constructed from its own inputs, a moderate-circularity score of 6 is warranted.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The framework modules are conceptual, not new physical entities. The main assumptions are about the completeness of the literature review, the adequacy of the small expert interview sample, the validity of survey ratings as evidence, and the decomposition of social compliance into modules. No free parameters or invented entities are present.

assumptions (4)
  • ad hoc to paper Socially compliant driving can be decomposed into the modules of the proposed framework (socially compliant decision-making, safety constraint, trade-off management, bidirectional behavioral adaptation, spatial-temporal memory).
    This decomposition is the central organizing assumption of Section 3.2; it is presented as a synthesis of the review and interviews, not derived from a formal model.
  • domain assumption The 68 studies selected in the scoping review are representative of the state of the art in socially compliant AV research.
    The review depends on the completeness of the search and screening process described in Section 2.1, but no formal appraisal of study quality or coverage is performed.
  • domain assumption The ten informal expert interviews provide sufficient insight to identify critical research gaps.
    The interview sample is small and not systematically analyzed; the paper summarizes themes without a formal qualitative analysis method.
  • domain assumption Expert survey ratings on a Likert scale measure the actual importance and feasibility of the framework components.
    The survey validation in Section 4 assumes self-reported expert opinion is a valid proxy for technical necessity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Developing Socially Compliant Automated Vehicles: Advances, Expert Insights, and A Conceptual Framework." pith.science (2026). https://pith.science/paper/ADQV56PC

@misc{pith2026250106089,
  author       = {Pith},
  title        = {Pith review of: Towards Developing Socially Compliant Automated Vehicles: Advances, Expert Insights, and A Conceptual Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ADQV56PC}},
  note         = {Machine review of arXiv:2501.06089}
}
read the original abstract

Automated Vehicles (AVs) hold promise for revolutionizing transportation by improving road safety, traffic efficiency, and overall mobility. Despite the steady advancement in high-level AVs in recent years, the transition to full automation entails a period of mixed traffic, where AVs of varying automation levels coexist with human-driven vehicles (HDVs). Making AVs socially compliant and understood by human drivers is expected to improve the safety and efficiency of mixed traffic. Thus, ensuring AVs' compatibility with HDVs and social acceptance is crucial for their successful and seamless integration into mixed traffic. However, research in this critical area of developing Socially Compliant AVs (SCAVs) remains sparse. This study carries out the first comprehensive scoping review to assess the current state of the art in developing SCAVs, identifying key concepts, methodological approaches, and research gaps. An informal expert interview was also conducted to discuss the literature review results and identify critical research gaps and expectations towards SCAVs. Based on the scoping review and expert interview input, a conceptual framework is proposed for the development of SCAVs. The conceptual framework is evaluated using an online survey targeting researchers, technicians, policymakers, and other relevant professionals worldwide. The survey results provide valuable validation and insights, affirming the significance of the proposed conceptual framework in tackling the challenges of integrating AVs into mixed-traffic environments. Additionally, future research perspectives and suggestions are discussed, contributing to the research and development agenda of SCAVs.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 32 canonical work pages

  1. [7]

    In: 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pp

    Large scale interactive motion forecasting for autonomous driving: the waymo open motion dataset. In: 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 9690–9699. Fagnant, D.J., Kockelman, K.,

    work page 2021

  2. [9]

    IEEE Trans

    Human-like decision making for autonomous driving: a noncooperative game theoretic approach. IEEE Trans. Intell. Transport. Syst. 22, 2076–2087. Hirose, N., Shah, D., Sridhar, A., Levine, S.,

    work page

  3. [10]

    In: 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pp

    Social occlusion inference with vectorized representation for autonomous driving. In: 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pp. 2634–2639. Huang, Z., Liu, H., Wu, J., Lv, C., 2023a. Conditional predictive behavior planning with inverse reinforcement learning for human-like autonomous driving. IEEE Trans. Intell. Transp...

    work page 2023

  4. [12]

    In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Interpretable social anchors for human trajectory forecasting in crowds. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 15551 – 15561 . Krajewski, R., Bock, J., Kloeker, L., Eckstein, L.,

    work page 2021

  5. [13]

    In: 2018 21st International Conference on Intelligent Transportation Systems (ITSC), pp

    The highD dataset: a drone dataset of naturalistic vehicle trajectories on German highways for validation of highly automated driving systems. In: 2018 21st International Conference on Intelligent Transportation Systems (ITSC), pp. 2118 – 2125 . Krajewski, R., Moers, T., Bock, J., Vater, L., Eckstein, L.,

    work page 2018

  6. [14]

    In: 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), pp

    The rounD dataset: a drone dataset of road user trajectories at round abouts in Germany. In: 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), pp. 1 – 6 . Ł ach, Ł ., Svyetlichnyy, D.,

    work page 2020

  7. [15]

    In: 2018 21st International Conference on Intelligent Transportation Systems (ITSC), pp

    Microscopic traffic simulation using SUMO. In: 2018 21st International Conference on Intelligent Transportation Systems (ITSC), pp. 2575 – 2582 . Lu, H., Lu, C., Yu, Y., Xiong, G., Gong, J.,

    work page 2018

  8. [18]

    Computer Vision – ECCV 2016, 549 – 565

    Learning social etiquette: human trajectory understanding in crowded scenes. Computer Vision – ECCV 2016, 549 – 565 . Sadeghian, A., Kosaraju, V., Sadeghian, A., Hirose, N., Rezatofighi, H., Savarese, S.,

    work page 2016

Show all 34 references
  1. [19]

    In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    SoPhie: an attentive GAN for predicting paths compliant to social and physical constraints. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1349 – 1358 . SAE International,

    2019

  2. [20]

    Intention-aware autonomous driving decision- making in an uncontrolled intersection. Math. Probl Eng. 2016, 1025349 . Sun, L., Zhan, W., Chan, C.Y., Tomizuka, M.,

    2016

  3. [21]

    In: 2019 IEEE Intelligent Vehicles Symposium (IV), pp

    Behavior planning of autonomous cars with social perception. In: 2019 IEEE Intelligent Vehicles Symposium (IV), pp. 207 – 213 . Sun, L., Zhan, W., Tomizuka, M., Dragan, A.D.,

    2019

  4. [22]

    Transport

    Influence of connected and autonomous vehicles on traffic flow stability and throughput. Transport. Res. C Emerg. Technol. 71, 143 – 163 . Toghi, B., Valiente, R., Sadigh, D., Pedarsani, R., Fallah, Y.P., 2021a. Altruistic maneuver planning for cooperative autonomous vehicles ...

    2021

  5. [23]

    https:// doi.org/10.21949/1504477

    Next Generation Simulation (NGSIM) Vehicle Trajectories and Supporting Data. https:// doi.org/10.21949/1504477 . Valiente, R., Razzaghpour, M., Toghi, B., Shah, G., Fallah, Y.P.,

    doi

  6. [24]

    In: Ethnographic Praxis in Industry Conference Proceedings, pp

    Developing socially acceptable autonomous vehicles. In: Ethnographic Praxis in Industry Conference Proceedings, pp. 522 – 534 . Y. Dong et al. Communications in Transportation Research 5 (2025) 100207 22 Wang, B., Su, R., Huang, L., Lu, Y., Zhao, N., 2024a. Distributed coopera...

    2025

  7. [25]

    In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    PANDA: a gigapixel-level human-centric video dataset. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3268 – 3278 . Wang, Z., Gao, P., He, Z., Zhao, L., 2021b. A CGAN-based model for human-like driving decision making. In: 2021 IEEE Wireless...

    2020

  8. [26]

    https://doi.org/10.48550/arXiv.2301.00493

    Argoverse 2: next Generation Datasets for self-driving Perception and Forecasting. https://doi.org/10.48550/arXiv.2301.00493 . Xia, C., Xing, M., He, S.,

    doi arXiv

  9. [27]

    In: 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), pp

    To develop human-like automated driving strategy based on cognitive construction: appraisal and perspective. In: 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), pp. 1 – 8 . Xu, C., Zhao, W., Wang, C., Cui, T., Lv, C.,

    2020

  10. [28]

    IEEE Trans

    Driving behavior modeling and characteristic learning for human-like decision-making in highway. IEEE Trans. Intell. Veh. 8, 1994 – 2005 . Xu, Y., Shao, W., Li, J., Yang, K., Wang, W., Huang, H., et al.,

    1994

  11. [29]

    In: 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC), pp

    SIND: a drone dataset at signalized intersection in China. In: 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC), pp. 2471 – 2478 . Xue, J., Zhang, D., Xiong, R., Wang, Y., Liu, E.,

    2022

  12. [30]

    In: 2019 American Control Conference (ACC), pp

    Social force aggregation control for autonomous driving with connected preview. In: 2019 American Control Conference (ACC), pp. 1388 – 1393 . Zhan, W., Sun, L., Wang, D., Shi, H., Clausse, A., Naumann, M., et al.,

    2019

  13. [31]

    https://github.com/alibaba-damo-academy/ universe

    Universe Simulator. https://github.com/alibaba-damo-academy/ universe . Zhang, L., Dong, Y., Farah, H., van Arem, B., 2023a. Social-aware planning and control for automated vehicles based on driving risk field and model predictive contouring control: driving through round abou...

    2023

  14. [32]

    In: Conference on Robot Learning 2021, pp

    SMARTS: scalable multiagent reinforcement learning training school for autonomous driving. In: Conference on Robot Learning 2021, pp. 264 – 285 . Zhu, M., Wang, Y., Pu, Z., Hu, J., Wang, X., Ke, R.,

    2021

  15. [33]

    In: 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), pp

    Human-like decision making and planning for autonomous driving with reinforcement learning. In: 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), pp. 3922 – 3929 . Yongqi Dong received the B.S. degree in telecommunication engineering from Be...

    2023

  16. [2009]

    In: 2009 IEEE 12th International Conference on Computer Vision, pp

    You ’ ll never walk alone: modeling social behavior for multi-target tracking. In: 2009 IEEE 12th International Conference on Computer Vision, pp. 261 – 268 . Peng, Z., Li, Q., Hui, K.M., Liu, C., Zhou, B.,

    2009

  17. [2011]

    Driving Assessment Conference 6, 2–9

    Fully automated driving: the road to future vehicles. Driving Assessment Conference 6, 2–9. Y. Dong et al. Communications in Transportation Research 5 (2025) 100207 21 Joo, Y.K., Kim, B.,

    2025

  18. [2016]

    In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Social LSTM: human trajectory prediction in crowded spaces. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 961–971. Arksey, H., O'Malley, L.,

    2016

  19. [2017]

    In: Conference on Robot Learning 2017, pp

    CARLA: an open urban driving simulator. In: Conference on Robot Learning 2017, pp. 1–16. Du, Y., Chen, J., Zhao, C., Liu, C., Liao, F., Chan, C.Y.,

    2017

  20. [2018]

    In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Social GAN: socially acceptable trajectories with generative adversarial networks. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 2255–2264. Hang, P., Huang, C., Hu, Z., Lv, C., 2022a. Decision making for connected automated vehicles at urban inte...

    2018

  21. [2019]

    In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Argoverse: 3D tracking and forecasting with rich maps. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8740–8749. Chang, W.J., Tang, C., Li, C., Hu, Y., Tomizuka, M., Zhan, W.,

    2019

  22. [2020]

    In: 2020 IEEE Intelligent Vehicles Symposium (IV), pp

    The InD dataset: a drone dataset of naturalistic road user trajectories at German intersections. In: 2020 IEEE Intelligent Vehicles Symposium (IV), pp. 1929–1934. Brown, B., Broth, M., Vinkhuyzen, E.,

    2020

  23. [2021]

    In: Proceedings of the 2021 IEEE International Conference on Human‒Machine Systems, pp

    Human-centric autonomous driving in an AV-pedestrian interactive environment using SVO. In: Proceedings of the 2021 IEEE International Conference on Human‒Machine Systems, pp. 1–6. Da, L., Wei, H.,

    2021

  24. [2022]

    In: 2022 IEEE Intelligent Vehicles Symposium (IV), pp

    The exiD dataset: a real-world trajectory dataset of highly interactive highway scenarios in Germany. In: 2022 IEEE Intelligent Vehicles Symposium (IV), pp. 958 – 964 . Munn, Z., Peters, M.D.J., Stern, C., Tufanaru, C., McArthur, A., Aromataris, E.,

    2022

  25. [2023]

    In: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, pp

    The halting problem: video analysis of self- driving cars in traffic. In: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, pp. 1–14. Buckman, N., Pierson, A., Schwarting, W., Karaman, S., Rus, D.,

    2023

  26. [2025]

    His research interests include deep learning, transportation big data, auto - mated driving, and traffic safety

    He is currently working as a researcher and group leader with the Institute of Highway Engineering, RWTH Aachen University. His research interests include deep learning, transportation big data, auto - mated driving, and traffic safety. He sought to employ artificial intellige...

    2025

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.