HUMANOID safety TECHNICAL TALK
FROM VIRTUAL AI TO PHYSICAL STABILITY.
Beyond Safe Stop — designing Safe Stability for humanoids operating around people.
With:
Dr.-Ing. Florian Weißhardt
Dr.-Ing. Tim Fröhlich (Synapticon)
Live + Q&A
Limited seats
YOUR POLICY WORKS IN SIM. THE ROBOT FAILS IN THE BODY.
Humanoids don’t break because the neural net is “wrong.”
They break because the system can’t guarantee timing, coordination, and stability across 50+ DoF in the real world.
They break because the system can’t guarantee timing, coordination, and stability across 50+ DoF in the real world.
Hard real-time control
Deterministic guarantees across 50+ DoF
Hardware integration that scales
Actuators + drives + FSoE backbone
“Brain vs Body” best practices
Protecting the stability loop
System-level integration
AI, motion and safety aligned
WHAT YOU WILL GAIN
Real-time boundaries
What must never be left to probabilistic logic.
System architecture patterns
How motion, perception and safety layers interact.
Safety transition strategies
Reducing torque without destabilizing system.
A reference architecture
To evaluate your current design decisions.
Live Q&A included
Technical brief provided after the session.
BUILT FOR HUMANOID BUILDERS
This session is designed for engineers who own system-level architecture decisions:
- CTO
- Robotics Architect
- Controls Lead
- Humanoid System Owner
If you are responsible for the path from prototype to reliable motion — this is for you. Humanoid stability is not a feature. It is an architectural decision.
