PARA
|
This projects develops plan-based control mechanisms for human-robot interaction, where the robot assists the person in everyday tasks and adapts to the person's abilities, expectations and preferences. The joint human-robot plan is represented explicitly in the robot program. To opimize the joint execution, we develop methods for representing and learning models of a person's capabilities, expectations and preferences. We apply this research in different domains, especially in the context of elderly care, where a robot assistant can enable elderly people to live independent, who currently depend on the help of others.
|
|
Application Domain
- Multi-Joint Action
- Intelligent Kitchen
Project Details
Robot technology is moving more and more into everyday environments, where the presence of humans is an additional level of complexity in robot control. Future robots will interact with people directly and assist them in everyday tasks. This requires a mutual understanding of humans and robots for coordinating their activities.
We extend a plan-based control language to represent joint human-robot plans in the program. In this way, tasks are assigned to the robot or the human or can be marked as a joint activity. Our framework allows to assign tasks on all levels of abstraction from high-level plans like setting the table to single action like moving an object. With this hierarchical planning approach, joint tasks can be controlled very flexibly. For assigning a task to the robot or the human, the robot needs an understanding of feasible actions of both partners as well as social rules and personal preferences of the person. Explicit models of these aspects will be learned by the robot using its own observations. Besides, we will develop general models for social interaction in collaboration with psychologists.
Current research on human-robot interaction is mostly focused on lower-level manipulation and communication activities. Planning approaches usually consider humans as an additional constraint on the robot plan or generate human-robot plans on a very high level of abstraction. But deciding on collaboration at the abstract planning level is not flexible enough to adapt the robot behavior to the current situation and user. In our approach, the robot has knowledge about the person and uses it to produce understandable, flexible joint plans that are appropriate in the situation at hand.
Joint human-robot scenarios are especially interesting for assisting special user groups, for example elderly people. Currently, these people have to rely on the help of other people, which is not always available and can be very expensive. With a robot assistant that interacts with people and helps them in their daily routine, people could live more a more independent life.
This project is partly funded by CoTeSys and BFHZ-CCUFB.
Material
- Posters
- Planning for Adaptive Robot Assistance (CoTeSys Fall Workshop 2009)
- Planning for Adaptive Robot Assistance (CoTeSys Spring Workshop 2009)
- Presentations
- A Planning and Learning Framework for HRI (ICAPS Workshop on Planning and Plan Execution for Real-world Systems, 2009)
- Planning for Adaptive Robot Assistance (2009)
- Videos
- Simulated human-robot collaboration in household activity
An autonomous robot and a human-controlled robot jointly setting the table. The red robot is controlled by a human by the keyboard (as in a computer game), the blue robot uses a control program. - Helping in a robot-robot task
It is the job of the blue robot to bring a plate to the table. The red robot observes if the blue one succeeds in the task. In the section trial, the blue robot moves away to the living room. Observing this, the red robot takes over the blue robot's job.
- Simulated human-robot collaboration in household activity
Selected Publications
-
Dynamic Generation and Execution of Human Aware Navigation Plans, 2010, Proceedings of the Ninth International Conference on Autonomous Agents and Multiagent Systems (AAMAS),(BibTeX)
