Alexandra Kirsch

Human-aware robot navigation: A survey

Navigation is a basic skill for autonomous robots. In the last years human-robot interaction has become an important research field that spans all of the robot capabilities including perception, reasoning, learning, manipulation and navigation. For navigation, the presence of humans requires novel approaches that take into account the constraints of human comfort as well as social rules. Besides these constraints, putting robots among humans opens new interaction possibilities for robots, also for navigation tasks, such as robot guides. This paper provides a survey of existing approaches to human-aware navigation and offers a general classification scheme for the presented methods.

Legible robot navigation in the proximity of moving humans

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.

Exploiting human cooperation in human-centered robot navigation

Robot path planning has traditionally concentrated on collision-free paths. For robots that collaborate closely with humans, however, the situation is different in two respects: 1) the humans in the robots environment are not randomly moving objects, but cognitive beings who can deliberately make way for a robot to pass and 2) the quality of a navigation plan depends less on quantitative efficiency criteria, but rather on the acceptance of humans. In this paper, we introduce a robot navigation approach that takes into account human-centered requirements and the collaborative nature of the interaction between the human and the robot.

Strategies of locomotor collision avoidance.

Collision avoidance during locomotion can be achieved by a variety of strategies. While in some situations only a single trajectory will successfully avoid impact, in many cases several different strategies are possible. Locomotor experiments in the presence of static boundary conditions have suggested that the choice of an appropriate trajectory is based on a maximum-smoothness strategy. Here we analyzed locomotor trajectories of subjects avoiding collision with another human crossing their path orthogonally. In such a case, changing walking direction while keeping speed or keeping walking direction while changing speed would be two extremes of solving the problem. Our participants clearly favored changing their walking speed while keeping the path on a straight line between start and goal. To interpret this result, we calculated the costs of the chosen trajectories in terms of a smoothness-maximization criterion and simulated the trajectories with a computational model. Data analysis together with model simulation showed that the experimentally chosen trajectory to avoid collision with a moving human is not the optimally smooth solution. However, even though the trajectory is not globally smooth, it was still locally smooth. Modeling further confirmed that, in presence of the moving human, there is always a trajectory that would be smoother but would deviate from the straight line. We therefore conclude that the maximum smoothness strategy previously suggested for static environments no longer holds for locomotor path planning and execution in dynamically changing environments such as the one tested here.

Influence of legibility on perceived safety in a virtual human-robot path crossing task

In the future robots will more and more enter our daily life. If we want to increase their acceptance it is necessary that people feel safe in the surrounding of robots. As a prerequisite we think that the robots behavior has to be legible in order to achieve such a feeling of perceived safety. With our present experiment we assess the perceived safety participants feel when an autonomous robot is crossing their path. Therefore participants are presented with a video based scenario in first person perspective. The robot is moving with two different navigation algorithms which allows us to test whether the legibility has an influence on the perceived safety and whether the two navigation algorithms differ regarding their resulting legibility and thus perceived safety. Results show that legibility as defined here increases perceived safety of both navigation methods while the level of perceived safety differs between them.

Plan-based Control of Joint Human-Robot Activities

Cognition in technical systems is especially relevant for the interaction with humans. We present a newly emerging application for autonomous robots: companion robots that are not merely machines performing tasks for humans, but assistants that achieve joint goals with humans. This collaborative aspect entails specific challenges for AI and robotics. In this article, we describe several planning and action-related problems for human-robot collaboration and point out the challenges to implement cognitive robot assistants.

RPL/sub LEARN/: extending an autonomous robot control language to perform experience-based learning

In this paper, we extend the autonomous robot control and plan language RPL with constructs for specifying experiences, control tasks, learning systems and their parameterization, and exploration strategies. Using these constructs, the learning problems can be represented explicitly and transparently and become executable. With the extended language we rationally reconstruct parts of the AGILO autonomous robot soccer controllers and show the feasibility and advantages of our approach.

Human-robot interaction in the MORSE simulator

Over the last two years, the Modular OpenRobots Simulation Engine (MORSE) project1 went from a simple extension plugged on the Blenders Game Engine to a full-fledged simulation environment for academic robotics. Driven by the requirements of several of its developers, tools dedicated to Human-Robot interaction simulation have taken a prominent place in the project. This late breaking report discusses some of the recent additions in this domain, including the immersive experience provided by the integration of the Kinect device as input controller. We also give an overview of the experiences we plan to complete in the coming months.

Robot learning language - Integrating programming and learning for cognitive systems

One central property of cognitive systems is the ability to learn and to improve continually. We present a robot control language that combines programming and learning in order to make learning executable in the normal robot program. The language constructs of our learning language RoLL rely on the concept of hierarchical hybrid automata to enable a declarative, explicit specification of learning problems. Using the example of an autonomous household robot, we point out some instances where learning-and especially continued learning-makes the robot control program more cognitive.

Evaluating directional cost models in navigation

A common approach to social distancing in robot navigation are spatial cost functions around humans that cause the robot to prefer paths that do not come too close to humans. However, in unpredictably dynamic scenarios, following such paths may produce robot behavior that appears confused. The concept of directional costs in cost functions [9] is supposed to alleviate this problem without incurring the problem of combinatorial explosions using temporal planning. With directional cost functions, a robot attempts to solve spatial conflicts by adjusting the velocity instead of the path, where possible. To complement results from simulations, in this paper we describe a user study we conducted with a PR2 robot and human participants to evaluate the new cost function type. The study shows that the real robot behavior is similar to the observations in simulation, and that participants rate the robot behavior less confusing with the adapted cost model. The study also shows other important behavior cues that can influence motion legibility. Categories and Subject Descriptors H.4 [Information Systems Applications]: Miscellaneous; D.2.8 [Software Engineering]: Metrics—complexity measures, performance measures