Jump to: Research Videos
Cognitive Robotics is concerned with endowing robotic or software agents with higher level cognitive functions that involve reasoning, for example, about goals, perception, actions, the mental states of other agents, collaborative task execution. Our research has mainly been on bridging the gap between high-level reasoning and low-level control, involving both theoretical and hands-on components.
Cognitive Factories Cognitive factories are a new paradigm in production engineering towards a more flexible, adaptable, and reliable production. According to this paradigm, the machines and processes in a factory are equipped with cognitive capabilities that involve reasoning about goals, perception, actions, collaborative task execution, etc. to allow them to assess and increase their scope of operation autonomously.
We propose to use causality-based formal representation and automated reasoning methods from artificial intelligence for such a cognitive factory with multiple teams of robots and humans, where each team tries to complete a complex assembly task, and where teams communicate with each other for efficiently sharing common resources.
Integrating Diagnostic Reasoning in Execution Monitoring For reliable and fault tolerant operation of cognitive factories, we introduce an algorithm to monitor plan executions. According to this algorithm, when some changes or discrepancies are detected, appropriate decisions are given based on the causes of these changes or discrepancies. To identify these causes (e.g., broken robots or robot components), we introduce a novel diagnostic reasoning method which synergistically integrates hypothetical reasoning, geometric reasoning, and learning from earlier experiences. Based on these causes, if necessary, new hybrid plans (task plans integrated with feasibility checks) are computed to reach the manufacturing goals by allowing repairs of robots/components. Below we provide a video of dynamic simulation of our execution monitoring algorithm with Kuka youBots and a Nao humanoid robot.
Ontological Reasoning for Rehabilitation Robotics Physical rehabilitation therapy is indispensable for treating neurological disabilities. Using robotic devices to assist repetitive and labor intensive rehabilitation exercises help decrease physical burden of the therapists and application related costs. As the number of rehabilitation robots increase, the information about them also increases, but most of the time in unstructured forms (e.g., as text in publications). This makes it harder to access the requested knowledge and thus reason about it. To facilitate access to requested knowledge about rehabilitation robots, we have designed and developed the first formal rehabilitation robotics ontology, called RehabRobo-Onto, in OWL, collaborating with experts in robotics and in physical medicine. To facilitate easy modification and use of RehabRobo-Onto by experts, we have developed a software (called RehabRobo-Query) with an intelligent user-interface. In this way, experts do not need to know logic-based representation languages of ontologies, like OWL, or Semantic Web technologies, for information entry, retrieval and modification. The ontology system consisting of RehabRobo-Onto and RehabRobo-Query is of great value to robot designers as well as physical therapists and medical doctors. On the one hand, robot designers can benefit from the system, for instance, to identify robotic devices targeting similar therapeutic exercises or to determine systems using a particular kind of actuation-transmission pair to achieve a range of motion that exceeds some threshold. Availability of such information may help inspire new designs or may lead to a better decision making process. On the other hand, physical therapists and medical doctors can utilize the ontology to compare rehabilitation robots and to identify the ones that serve best to cover their needs, or to evaluate the effects of various devices for targeted joint exercises on patients with specific disorders.
Multi-Agent Path Planning Many tasks, such as computer games, street sweeping, mail delivery, and robotic surveillance and patrol, vehicle routing, environmental monitoring, require multiple agents to visit points in an environment to accomplish a goal, ensuring that they do not collide with static obstacles or other moving agents. We study such problems in a general framework using high-level representation formalism and efficient solvers of the declarative programming paradigm Answer Set Programming.