Hide menu

The Knowledge Processing Laboratory

The Knowledge Processing Laboratory (KPLAB) was established in 1996. Professor Patrick Doherty heads the lab. There are currently 2 professors, 2 research assistants and 6 PhD students.

Research in KPLAB focuses on the theoretical and practical aspects associated with the representation of knowledge and the reasoning and inference techniques associated with the processing of knowledge as used by both physical and software artifacts.

KPLAB also includes three topic-focussed research groups:

  • Planning and Diagnosis
  • Cognitive Robotics
  • Applied Logic

Research topics of current interest include the following:

  • Multi-Agent Systems - Research with multi-agent systems involves the study and development of AI problem solving and control paradigms for both single and multi-agent systems where issues related to interaction, cooperation, autonomy, and distribution are paramount.
  • Planning and Diagonosis - Research with automated planning involves the study and development of algorithms which generate strategies or sequences of actions to achieve goals. Research with automated diagnosis involves the study and development of algorithms which capitalize on cause-effect information in a system or environment of a system in order to trouble shoot and provide explanations and remedies for faulty system or cognitive behavior.
  • Cognitive Robotics - Research with cognitive robotics involves the study and development of higher-level cognitive functions that involve reasoning and empirically testing such functions on deployed robotic systems. Central to the endeavour is the efficient use and and representation of models of both the robot and its embedding environment and grounding these models in such environments through sensing and perception systems.. Logic is often the modeling language of choice in this respect.
  • Applied Logic - Research in applied logic involves the study and use of logic as a representational mechanism for constructing models and a reasoning mechanism for using such models in intelligent artifacts such as software agents or robotic systems.

Special emphasis is currently being placed on the design and development of command and control architectures for unmanned aerial vehicles (UAVs) and their integration with active vision systems and other sensors. Such systems require on-line planners, prediction and chronicle recognition mechanisms, GIS and soft real-time databases, and a variety of knowledge representation frameworks with associated inference mechanisms used to dynamically construct and reason about the UAV's internal and external environment.

Page responsible: Patrick Doherty
Last updated: 2014-04-30