This work group focusses on perception, reasoning, prediction, and motor action in a simple closed-loop system. Use your own computing hardware, or use our neuromorphic SpiNNaker or GPU setup.
We provide small mobile robots, equipped with high-resolution event camera and some additional sensors. We setup a soft-ball on a string as swinging pendulum (obstacle). Your robot shall go by the pendulum without getting hit. We expect your robot brain will need to understand how the pendulum swings, and predict into the (near) future for a safe pass-by. Maybe you find a better solution?
It's all about perception, reasoning, prediction and motor action in closed-loop systems :)
We hope to see you!
Reference implementation for control available on GitHub: https://github.com/ncskth/durin/
Recordings for training are available at the following (local) URLs:
Static (standing) 10s recording with 5ms time windows: http://10.1.18.67:8000/standing.dat
Moving (slowly) 10s recording with 5ms time windows: http://10.1.18.67:8000/moving.dat
PLEASE NOTE that you need to be on our WiFi access point to operate the robots.
|Tue, 03.05.2022||15:30 - 16:00||Disco (lab)|