Safe Physical Intelligence
About Us
SafePI.ai works on making robotic systems safer by covering different protection aspects.
You can learn more about our work and follow updates on our LinkedIn page
-
Unpacking Hiwonder SO-ARM101
The Hiwonder SO-ARM101 provides an ideal platform for an initial implementation and validation of our robotics safety framework. It is accessible and low-cost, has clearly defined joint structure and straightforward Python-based control.
-
Project status report - September'26
The video report is available on our YouTube Chanel
-
🌱 Impact Social Cup in Bilbao!
We were delighted to take part in the Bilbao stop of the Impact Social Cup, an initiative that promotes impact-driven entrepreneurship and brings together startups, investors, and key players from the entrepreneurial ecosystem.
-
Safe Physical AI in Urban Mobility
We’re happy to share our latest paper, “Ensuring Safe Physical AI in Urban Mobility via Hazard-Informed Synthesized Envelopes”, to be presented at CRET 2026.
-
SafePi Presented at ICAME 2026 in Barcelona
On 14 July 2026, Rostislav Yavorskiy, CTO of SafePi, presented our project at the 4th International Conference on Automation and Mechatronics Engineering, Barcelona, Spain, ICAME’26.
-
Learning of Robot Safety Policies via Synthetic Scenarios
We’re happy to share our latest report, “Learning of Robot Safety Policies via Adversarial Synthetic Scenarios”.
-
Our project to be presented at the International Conference on Automation and Mechatronics Engineering
Our paper has been accepted for presentation at the International Conference on Automation and Mechatronics Engineering (ICAME’2026).
-
Safety layers
Here is a quick look at the essential layers of protection:
-
We need machines that...
We need machines that belong in our world.