What is an agent: agents and objects; autonomous decision making; typical application areas for agent systems. Abstract architectures for agents; tasks for agents; the design of intelligent agents - reasoning agents, agents as reactive systems (e.g, subsumption architecture); hybrid agents (e.g, PRS); layered agents (e.g, Interrap) The sense - decide - act loop. Sensors: passive versus active sensors; light sensors; infra-red sensors; ultrasound sensors. Actuators: motors & servo motors; kinematics; manipulators. Movement: path planning; localisation; Principles of SLAM (Simultaneous Localisation and Mapping), including Bayesian Beliefs, Kalman Filters, Probablistic Sensor Models and Probablistic Motion Models. A contemporary experimental robotics platform. Guest lectures covering contemporary topics in Robotics will also be delivered. The schedule of topics is as follows: - Introduction to Robotics, and the Development API - Wheeled based Kinem
atics, Locomotion & Odometry - Beliefs and Bayesian Filters - Agents and Behaviour Based Robots - Probabilistic Motion Model - Advanced Perception and Probabilistic Sensor Model - Markov Localisation and Particle Filters - Maps, Landmarks and Mapping with Known Poses - Kalman Filters, Simultaneous Localisation and Mapping (SLAM) - Exploration, Navigation and Obstacle Avoidance
|