HUMANOID EXPERTTALK
Why "Safe Stop" fails and what replaces it.
With Florian Weißhardt, Ph.D., and Tim Fröhlich, Ph.D. (Synapticon).
Safety transitions can destabilize a humanoid.
Cut torque at the wrong moment and you create instability — not protection.
Deterministic multi-axis control
Hard real-time guarantees across 50+ degrees of freedom
Modeling volumetric safety
Dynamic 3D fall zone awareness
AI behavior supervision
Constraining non-deterministic decisions
Integration at the system level
Drives, perception, and safety as one architecture
WHAT YOU WILL GAIN
A System Blueprint
What must remain deterministic and why.
An architectural model
How motion, supervision, and safety layers interact.
A transition strategy
Designing safe torque reduction.
Checkliste für die Bereitstellung
From demo to real-world operation.
BUILT FOR HUMANOID SYSTEM OWNERS
This session is designed for anyone who defines system-level architecture decisions.
- chief technology officer
- Head of Robotics
- Chief Engineer
- Controls Lead
- Safety & Architecture Owner
Humanoid safety is not a feature decision. It is an architecture decision.
