Model Validation: Robot Behavior in People Guidance Mission using DTM model and Estimation of Human Motion Behavior. Ana´ ıs Garrell and Alberto Sanfeliu Institut de Rob` otica i Inform` atica Industrial (CSIC-UPC) 08028 Barcelona, Spain Abstract— This paper describes the validation process of a simulation model that have been used to explore the new possibilities of interaction when humans are guided by teams of robots that work cooperatively in urban areas. The set of experiments, which have been recorded as video sequences, show a group of people being guided by a team of three people (who play the role of the guide robots). The model used in the simulation process is called Discrete Time Motion model (DTM) described in [7], where the environment is modeled using a set of potential fields, and people’s motion is represented through tension functions. The video sequences were recorded in an urban space of 10.000 m 2 denominated Barcelona Robot Lab, where people move in the urban space following diverse trajectories. The motion (pose and velocity) of people and robots extracted from the video sequences were compared against the predictions of the DTM model. Finally, we checked the proper functioning of the model by studying the position error differences of the recorded and simulated sequences. I. INTRODUCTION Nowadays, research community have increased its interest in robots performing social tasks, consequently the needed of studying new abilities for human-robot interaction and task cooperation have become important issues. One application where these abilities are shown is in guiding a group of people using one or several robots [4]. Some experiments of guiding people with one robot have been already performed, but when is a group of people guided by a team of robots, the experiments are not performed easily due to the limitations of the present technology (robots do not move or change their orientation fast enough). In these situations is better to simulate the experiments and then validate the results using sequence of real guided missions [2]. In this work, we present the validation of a model described previously in [7], denominated “Discrete-Time- Motion” model (DTM). The DTM model is a new approach for guiding people in open areas of urban settings using multiple robots acting in a cooperative way. One of the robots is the leader, and acts as a human tour-guide. It is placed at the front of the formation and its role is to estimate the trajectory of both people and the rest of robots. The other robots, called shepherd, are responsible for guiding people, This research was conducted at the Institut de Rob` otica i Inform` atica Industrial (CSIC-UPC). It was partially supported by CICYT projects DPI2007-61452 and Ingenio Consolider CSD2007-018, by CSIC project 200850I055 and by IST-045062 of European Community. The first author acknowledges Spanish FPU grant ref. AP2006-00825. Fig. 1. (a), (b) and (c) show different captions of the guided group by the camera network. (d) shows an aerial view of the experimental site, Barcelona Robot Lab, including the distribution of camera network, and the trajectory of the group in one of the experiments. preventing any person leaving the group, and following the path given by the leader. The DTM model represents robots, people and obstacles by means of potential functions, and it computes the estimations of people’s and robots positions. The model uses a particle filter [1] for positions estimation, and it considers Helbing equations [8], [9] for the simulation of human reactions in front of different stimulus that come from the robots and the constraints of the environment. The works presented in the literature show their contri- butions for groups of robots that interact with people using simulations. For instance, [6] performs a qualitative analysis of the movements of different entities –such as humans or animals– and build an architecture of three robots to guide them. However, realistic situations, such as the existence of obstacles or dealing with people leaving the group are not considered, and their models were not validated. In [13] several types of robot formations and different strategies for approaching the robots to people are considered. However. all these issues and the general movements of the robots are ruled by a large number of heuristics which makes the system impractical. Furthermore, in order to achieve the The 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems October 18-22, 2010, Taipei, Taiwan 978-1-4244-6676-4/10/$25.00 ©2010 IEEE 5836