The PRASYS team – Perception, Robotics and Autonomous Systems

Such systems are based on three pillars: perception, action and the interaction between the two. Any autonomous perception system must be able to simultaneously analyze the data collected in relation to its immediate environment, become familiar with this environment and, depending on what this environment is, autonomously organize the steering and navigational aspects of its mission. This recent issue stems from the field of robotics in which the robot-environment interaction occurs via the sensor, which requires a study of the action-perception mechanisms.

Two main areas of focus have thus been defined: perception and action. Interaction between these two areas of focus is then essential for the system to be autonomous.

•  Research focus 1: Autonomous action
•  Research focus 2: Autonomous perception

Perception (research focus 2) is the process by which the system attempts to understand its observations and devise a relevant representation, not only of the environment’s current state, but also of its uncertainty. Using the perception’s results, the system must then reason out the actions (research focus 1) it must implement to achieve its mission goals, according to certain criteria such as efficiency, safety, speed, and energy consumption.

APPLICATIONS

Defense and security, underwater archaeology, harsh environment operations

EXPERTISE

The team comprises researchers with expertise in information processing, robotics and automation. The team provides new mathematical tools, including set theory methods, suitable for managing uncertainties in unstructured environments such as the marine environment.