The effect of appearance features of sidewalk autonomous delivery robot on perception of pedestrian

Sidewalk autonomous delivery robots (SADRs) are increasingly appearing on urban sidewalks and sharing public space with pedestrians. This study aims to investigate how different appearance features of SADRs influence four dimensions of human perception: Likeability, Competence, Perceived Safety, and Existence Acceptance. In Study 1, an affinity diagram method was used to classify the appearance features of existing SADRs into four categories—contour, proportion, facial features, and color scheme—which were then used as the design variables in Study 2.

In Study 2, 100 participants completed an online questionnaire after viewing AI-generated videos of SADRs and evaluated the robots on the four perceptual dimensions. The statistical results showed that rounded contours and the presence of facial features significantly increased Likeability, Perceived Safety, and Existence Acceptance. White–black and yellow color schemes generally received higher ratings for Likeability, Perceived Safety, and Existence Acceptance than black–gray and red color schemes.

In addition, a tall proportion significantly reduced Likeability. Study 3 used online semi-structured interviews to further examine the reasons underlying these evaluations and to explore participants’ subjective perceptions of SADRs. The findings provide practical guidance for designing SADR appearances that can foster more positive pedestrian perceptions and acceptance.

This work was supported by grants awarded to Fan Xu from the Ningbo Philosophy and Social Science Laboratory Project, and the Scientific Research Foundation of NBUT (Grant No. 24KQ086), and the Ningbo Education Science Planning Project (Grant No. 2026YGH042). Robotics Institute, Ningbo University of Technology, Ningbo, 315211, China College of Design, Hanyang University, Ansan, 15588, Republic of Korea College of Art and Communication, China Jiliang University, Hangzhou, 310018, China Design College, Yunnan Arts University, Kunming, China College of Mechanical Engineering, North China University of Science and Technology, Tangshan, Hebei, China The authors declare no competing interests. This research project has been reviewed and approved by the Robotics Institute of Ningbo University of Technology, which serves as the local approval body in the area where the research was conducted.

The Institute, a secondary-level body, conducted the ethical review as its parent institution does not have a centralized Institutional Review Board (IRB). Prior to the online questionnaire surveys of the study, all participants will be provided with a comprehensive information sheet detailing the research purpose, ethical review status, principles of data confidentiality and anonymity, and the contact information of the research team. Participants will only proceed to the experimental phase after they have fully understood and signed the informed consent form.

All procedures followed in this study adhere to the ethical standards of the Robotics Institute of Ningbo University of Technology. Robust measures are in place to ensure the creation of a completely anonymized dataset, thereby safeguarding participant privacy and confidentiality throughout the research process. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.