Arm's O'Driscoll to talk "Building Physical AI that Scales" at RoboBusiness

Arm's framing starts from recent advances in foundation models, vision-language-action systems, and edge AI, which the company says are expanding what robots can sense, reason about, and do. The next challenge, in Arm's view, is turning those capabilities into autonomous systems people can trust in the real world — whether restoring mobility through intelligent prosthetics, delivering critical medical supplies, or helping workers operate more safely alongside intelligent machines.

As robots become more capable, engineering priorities shift toward system-level design. Arm points to how intelligence is distributed across the robot, how learned and deterministic behaviors interact, and how compute platforms balance real-time responsiveness, efficiency, and safety. Arm's recently released Total Design, which brings together expertise across the technology stack to reduce integration complexity and accelerate autonomous system development, sits in that same direction.

O'Driscoll's session — drawing on lessons from across the robotics ecosystem — is positioned as a practical framework rather than a product announcement. The value for attendees likely hinges on whether the compute principles he presents translate into decisions they can apply as they scale to production, particularly around safety and reliability. Any broader shift in how the industry approaches physical AI would depend on whether such frameworks are actually adopted by development teams.