Perception for autonomous systems

Perception of autonomous systems System TRL Levels: 3-7 Linked WPs: 3,7 Linked UCs: 1, 3
AI, Deep Learning, Autonomous vehicles, Cloud
Automotive, Robotics Potential market entry: 2022 Re-engineering effort: 15-30 MM
Since the dawn of the AI in the automotive industry, studies follow the human behavior which is made of: Perception, Understanding, Decision and Act (PUDA). The autonomous vehicle respects exactly the same behavior and in our case, a computer has to reproduce this behavior.

Researches in this domain is wide and we need to seal the basics from the Perception. That is why we expect to create this first PIARCH Perception to let the next generation build the others.

For this PIARCH Perception, we intend to select all the available sensors that a vehicle can get. As the 5 human senses, if we have all both, the better will be the perception else we still be able to compensate to have a trustful perception for the rest of the chain of action: PUDA. This PIARCH expects to manage from one to several raw sensors to a scene understanding.