A pool of water as a reservoir computer hits 100% accuracy on robotic obstacle avoidance
bravo_abad · x · 2026-09-11
Researchers used a pool of water as a reservoir computer: six actuators generate waves whose interference turns six distance-sensor readings into a high-dimensional physical representation for robotic obstacle avoidance.
- Key finding: input timing must match the reservoir's natural dynamics — waves too close produce similar patterns; too far apart, they barely interact.
- For this system, accuracy peaks with 200 ms spacing between adjacent inputs; reading one interference pattern 150 ms later yields up to 100% obstacle recognition accuracy.
The authors argue physical AI performance depends not only on models and hardware but on how information is injected into the physics itself.
More from Embodied
- TARS demos its AI World Engine: robots carrying learned skills to novel tasks — socialwithaayan · 2026-09-11
- LIT trains VLAs to act before seeing: two-stage method fights visual shortcuts — DJiafei · 2026-09-11
- Latent Interface Training breaks vision-action shortcuts in VLA and WAM training — DJiafei · 2026-09-11
- Latent Interface Training boosts VLA generalization, lifting LIBERO-Plus success up to 11 points — DJiafei · 2026-09-11
- Robot nearly sticks the landing: training progress but leg lift still fails — hbouammar · 2026-09-11
- Robot's kick attempt goes from near-fall to almost sticking the landing — 0xSammy · 2026-09-11