AIICS

Patrick Doherty

LiU » IDA » AIICS » PatDo

Research

The major red thread in my research focuses on the theoretical foundations and the pragmatic engineering required to bring theory into practice in the development of intelligent autonomous systems. Current focus is placed on Unmanned Aircraft Systems (UAS:s), but the research should be considered generic to the extent that is possible. In other words, many results can be used with any complex physical or software-based autonomous system. Particular research areas of current interest are:

  • Knowledge Representation - I am interested in developing qualitative models for dynamic processes using logic as a representational tool. This work encompasses development of logics of action and change, non-monotonic reasoning, approximate reasoning and more recently integrating logical modeling with coalitional game theory. In addition, together with Fredrik Heintz, we have had an interest in attacking the sense-reasoning gap in autonomous systems through the use of stream-based reasoning components in robotic architectures. This work has recently been extended to the multi-agent case.
  • Automated Planning - My interest in planning originated with the development of a logic-based forward chainning planner based on the use of TAL together with Jonas Kvarnström. This idea has been extended through the years to include an execution monitoring component based on similar ideas which use tense logical formula progression. More recently, TALplanner has been extended for use in multi-agent scenarios. My interest here is to leverage this planning capability in collaborative scenarios where many robots are involved in solving particular tasks collaboratively.
  • Autonomous Unmanned Aircraft Systems - I am interested in developing highly autonomous UAS:s which can collaborate naturally with human operators and other robotic entities. This work ranges from low autonomy associated with traditional control technology to high autonomy associated with artificial intelligence technology. In particular, I am interested in integrating low- and high-autonomy in a robust and seamless manner. This implies developing novel architectural components in such systems and dealing pragmatically with the sense-reasoning gap and the symbol-grounding problem.
  • Multi-Agent Systems - The autonomous systems I am interested in rarely operate in isolation. They interact with other autonomous systems and with human resources. This observation should be kept in mind when developing robotic architectures to ensure robust operation in the larger multi-agent, multi-human environments. In this respect, I am interested in multi-agent architectures and mixed initiative probelm solving. This involves development of better models of delegation, collaboration, distributed planning, resource allocation and scheduling.

Much of the research I pursue is highly multi-disciplinary and consequently, highly collaborative and involves many colleagues and students in AIICS. For more details, please refer to our project descriptions and publications.

I currently participate in the following Swedish Excellence Centers:

  • [MOVIII] - Modelling, Visualization and Information Integration: A Center for Decision Support in Complex Systems.
  • [CADICS] - A Linnaeus Research Environment for Control, Autonomy, and Decision-making in Complex Systems.
  • [ELLIIT] - Excellence Center at Linköping - Lund for Information Technology.
  • [LINKLAB] - Linköping University - SAAB Excellence Center for Future Aviation Systems.