Abstract— Person-following is an important aspect in many service robotic applications whilst supporting a person in performing daily tasks. Few studies have actively worked towards making person-following behavior usable, pleasurable or personal. As such, user studies are essential for promoting the interaction design, and increase user satisfaction and acceptance. A specific experimental setup for studying of socially acceptable person-following preferences and algorithmic design is presented here. In six user studies (171 participants in total) following- related factors were examined, of those, two related to environmental influence are specified here. Objective and subjective measurement of the quality of the interaction and user satisfaction were taken. Results and implications are discussed. I. INTRODUCTION Person-following is an important aspect in many service robotic applications whilst supporting a person in performing daily tasks (e.g., carrying groceries, physical monitoring, and companionship). To create robots that move in socially acceptable manners it is important to consider a multitude of parameters such as the robots’ speed, acceleration and deceleration properties, the lead human’s walking speed, angle of following and the appropriate physical proximity, as a function of the environment, context, physical state and human intent. A recent in-depth review [1] identified four categories of factors that influence social considerations of person- following behaviors, including: the characteristics of the person being followed, the features of the robotic system, the task which is performed, and the environment. In order to design a socially aware person-following robot that meets user needs and desires, it is critical to consider the various factors that influence the way a person would like to be followed by a robot [2]. Still, the majority of published works on person-following robots focused on the technical aspects of making the person-following behavior safe and functional. Most applications set their velocity based on their proximity to an obstacle or to the person or based on specific environmental constraints like safe passage through a door. The distance maintained between the robot and the person it follows is often constant [1] although some implementations used Hall's human-human distancing [7] to select their following distances. This research was supported by the Ministry of Science, Technology & Space, Israel, Grant # 3-12060, “Follow me”, by the EU funded Innovative Training Network (ITN) in the Marie Skłodowska-Curie People Programme (Horizon2020): SOCRATES (Social Cognitive Robotics in a European Society training research network), grant agreement number: 721619 and partially supported by the Helmsley Charitable Trust through the Agricultural, Biological and Cognitive Robotics Center, the Marcus Endowment fund and by the Rabbi W. Gunther Plaut Chair in Manufacturing Engineering, George Figure 1. Experimental setup and dedicated person-following robot. The vast majority of research in person-following has focused on robots that follow people from behind, some work has been done on enabling robots to accompany people side- by-side or at an angle (e.g. [6]). In most studies, the relative position of the robot to the person remains fixed, and changes solely based on environmental considerations. The criteria for selecting the initial angle of following is often unspecified, especially when there is no task-specific reasoning. Very few studies have actively worked towards making the person- following behavior usable, pleasurable or personal. As such, user studies like ones specified here are essential for promoting interaction design, and increasing user satisfaction and acceptance. Furthermore, while relevant human-robot literature is sparse, research in human-human interaction reveals many environmental characteristics that affect distancing behavior [8]. People seem to allow others to come closer to them when they're in larger rooms with better lighting and higher ceilings. II. METHODS Experimental platform. A specific experimental setup for studying of socially acceptable person following preferences and algorithmic design is presented here (see Fig. 1) with an emphasis on the factors that were considered in six user studies, and in relation to the four categories (human-robot- task-environment). In this setup, a young adult was followed by a robot in various predefined use cases. For this, a Shrut Chair in Human Performance Management, all at Ben-Gurion University of the Negev. All authors are from the Department of Industrial Engineering and Management, Ben-Gurion University of the Negev Beersheba 84105, Israel (phone: +972-86461434; email: olatunji@post.bgu.ac.il, sarne@post.bgu.ac.il, shaneeh@post.bgu.ac.il, zaichyk@post.bgu.ac.il, tamarama@post.bgu.ac.il, orontal@bgu.ac.il, yael@bgu.ac.il) User preferences for socially acceptable person-following robots: environmental influence case studies Olatunji Samuel, Fleischmann-Serna Vardit, Honig S. Shanee, Zaichyk Hanan, Tamara Markovich, Oron-Gilad Tal, Member, Edan Yael, Member .