Legible robot navigation in the proximity of moving humans

Abstract

Our objective is to improve legibility of robot navigation behavior in the presence of moving humans. We examine a human-aware global navigation planner in a path crossing situation and assess the legibility of the resulting navigation behavior. We observe planning based on fixed social costs and static search spaces to perform badly in situations where robot and human move towards the same point. To find an improved cost model, we experimentally examine how humans deal with path crossing. Based on the results we provide a new way of calculating social costs with context dependent costs without increasing the search space. Our evaluation shows that a simulated robot using our new cost model moves more similar to humans. This shows how comparison of human and robot behavior can help with assessing and improving legibility.

Integrating human observer inferences into robot motion planning

“…If this is the only intent which explains the motion, the observer can immediately infer the actor's intent, meaning that the motion is also legible. This is why we tend to assume that predictability implies legibility -that if the robot moves in an expected way, then its intentions will automatically be clear [2,6,35].…”

“…Relation to Prior Work "Predictable" means "expected", and is usually used to refer to a desirable property of motion without needing additional clarification [2,6]. "Legible", on the other hand, is typically reserved for writing, and appears in the robot motion literature accompanied by an explanation: being legible means that observers are able to recognize [6] , infer [39], or understand [3,35] the intentions of the robot, or that robot indicates the goal it will reach [35] -overall, legibility is about expressing intent. Building on this, we define predictability and legibility as enabling inferences that the collaborator needs to make, and propose trajectory optimization criteria for the two.…”

Human-robot co-navigation using anticipatory indicators of human walking motion

“…A mixed-observability Markov decision process (MOMDP)-based model is used to reason about the intent and/or goals of human agents and plan around them in [8]. Studies concerning human factors issues include considerations of safety, proxemics and the legibility and predictability of robot motion [21]- [23].…”

Heuristic Decision-Making for Human-aware Navigation in Domestic Environments

“…The situation can change quickly, obstacles may not be on the map, either because they move themselves (like people or pets or other household robots) or because they are movable (like chairs). Relying too much on path planning in dynamic environments leads to illegible 1 behavior [11,14]. On the positive side, rooms are usually furnished in a way to make navigation easy for people and thus hardly contain dead ends.…”

“…It remains to be tested in dynamic environments, which allow for a much higher variance of situations. Typical navigation algorithms such as the ones implemented in ROS behave different in crossing situations than people [11,2] and are not really legible [14].…”