Arm executive to outline compute principles for trusted physical AI at RoboBusiness
Arm's VP of go-to-market for physical AI will speak at RoboBusiness about scaling physical AI systems. He will discuss distributing intelligence across cloud and edge, coordinating learned and deterministic behaviors, and ensuring safety and reliability. The talk aims to provide a framework for building trustworthy autonomous systems.
O’Driscoll’s session will center on how compute architectures must balance real-time responsiveness with efficiency and safety as robots move from prototypes to production. His background spans more than two decades at Arm, including roles in IT, EDA, and CPU and SoC design, which positions him to address both technical and commercial adoption challenges across emerging physical AI markets.
RoboBusiness, produced by The Robot Report, celebrates its 20th anniversary this year in Santa Clara. The event targets commercial robotics developers and covers applications in manufacturing, healthcare, agriculture, and logistics, with the full conference pass providing access to keynotes, technical sessions, and networking events.
The push toward trusted physical AI could reshape how autonomous systems are deployed in workplaces and daily life. If Arm’s framework gains traction, developers may build robots that more reliably assist in healthcare, logistics, and manufacturing, potentially improving worker safety and operational efficiency. However, the balance between learned and deterministic behaviors raises questions about accountability when systems fail, affecting regulators, insurers, and end users who must trust machines with critical tasks.