Keynote
In person
LeadersGoldDiscovery

Trustworthy robot teams in the real world

  • Date
    8 July 2026
    Timeframe
    11:00 - 11:30
    Duration
    30 minutes
    • Days
      Hours
      Min
      Sec

    As autonomous robots step out of the laboratory and into our streets, factories, and homes, trustworthy AI is no longer a digital abstraction, it is a physical necessity. This transition requires a fundamental shift from robots that simply “work” to robot teams that are probably safe, scalable, and accountable to the public. My talk begins with the foundation of mathematical safety, providing rigorous guarantees in dynamic environments and co-designing physical spaces to ensure robots and humans interact without incident. To bring these promises to the real world, we must ensure safety scales; we achieve this by combining the logic of traditional computing with the speed of modern AI, enabling agile, collision-free coordination for massive robot collectives. Finally, to solve the Black Box problem of AI behavior, I introduce the concept of “motion watermarking”. By embedding invisible digital signatures directly into a robot’s movement, we allow anyone with a smartphone to verify a robot’s behavior and intent in real-time, bridging the gap between physical action and digital provenance. Together, these pillars establish a technical and ethical framework for a future where autonomous teams are not just high-performing, but fundamentally responsible members of our global society.

    Are you sure you want to remove this speaker?