Inferring the goal of an approaching agent: A human-robot study

Abstract

Abstract-The ability to infer intentions and predict actions enables coordinating of one's own actions with those of another human and allows smooth and intuitive interaction. The aim to achieve equally effective human-robot interactions is a crucial aspect of current robotic studies. Thus, we assume that studying human-human interaction provides valuable insights allowing to implement mutual intention recognition and action prediction in robotic systems. A common scenario of interaction, be it in everyday life or in an industrial setting, is that two or more agents share the same workspace and perform tasks without interference. If humans are involved, the robots should act sufficiently predictable to enable the human to attribute goals and predict motion trajectories. In the present work, we first analyzed how well a human recognizes the goal of another person entering the room, and whether this ability is deteriorated by concealing gaze direction of the other person. In a second setup, the same experiment was repeated by replacing the approaching person with a wheeled robot. On average, the distance at which subjects predicted the goal of the approaching agent was approx. 4 m and depended on subject and goal position, but not on the type of agent. However, goal attribution showed a considerable proportion of errors for the robot (19%), much less for a human with hidden gaze direction (6%), and almost none for a human with visible gaze (1%). Thus, our subjects apparently decided on the goal of the approaching agent without taking into account the reliability of directional cues, thus resulting in more errors. In a human-robot setting, such wrong predictions about robotic behavior may easily lead to dangerous situations. For smooth and safe interaction, it is therefore important to ameliorate the predictability of robotic actions.

Influence of person- and situation-specific characteristics on collision avoidance behavior in human locomotion.

“…The observed stereotypical sequence of events is also compatible with a mechanism for social interactions suggested by Goffman (1971), who proposes a three-step process for interaction, comprising externalization (i.e., the choice of own actions so that the underlying intention can be recognized by others), scanning (i.e., the attempt to deduce others' intentions from their actions), and minimization of adjustment (i.e., the adaption of one's own plans to accommodate interference from nearby agents in such a way that adjustment with respect to the original plan is minimized). This is in line with our observations, namely that the person giving way waits for the other person to initiate a path adjustment and, upon seeing this, infers (Basili et al, 2012;Blake & Shiffrar, 2007) that the other person intends to cross first and, thus, adjusts their own trajectory to give way. By doing so, the risk of a conflicting role assignment is reduced and path adjustments can be minimized.…”

Human–Robot Interaction Video Sequencing Task (HRIVST) for Robot's Behavior Legibility

“…In summary, we found that the significant differences in participant scores for each video are consistent with the existing literature about the methods to achieve legible robot behaviors [15]. In particular, to generate legible robot behaviors, the following variables were proposed: human-like behavior ( [51], [52]); stereotypical motions of everyday activities (e.g., reaching for an object) ( [51], [53]); the addition of complementary gestures to clarify intentions (e.g., gaze, pointing) ( [54], [55]); and the greatest possible visibility of the robot motion ( [56], [57]). We agree that these variables reported may have played a role in the scores.…”

“I See What You Did There”

“…While predictability is an important concept to human-robot interactions, there is no common understanding about the concept of predictability and there are various ways to approach this concept-it is an umbrella concept. Predictability is used in terms of it being a property of robotic motion [4,23,24,55,61,93]. In this context, the predictability of a motion relates to how well it matches the person's prediction regarding the robot's motion trajectory, given that the goal of the robot is known [23].…”