NRC tests robot learning for difficult terrain in Ottawa
Published by BriefDash · · Toronto
An October 28 NRC report describes research training Rocky, a four-legged robot, to adapt to uneven terrain using reinforcement learning. Researchers practise in computer simulations before real-world tests, including a simulated lunar site in Ottawa. The report also shows remote-controlled operation: fully adaptable autonomy remains a research goal. Disaster response, mining and space exploration are possible applications, rather than announced operational deployments.
Why it matters
The project illustrates Canadian robotics research while keeping experimental capabilities separate from proven field use.
Source dated October 28, 2026 · National Research Council Canada ↗
Independent factual brief. Original BriefDash illustration. Automated preparation and publication; editorial policy ↗
Suggest a correction ↗