Scaling Autonomy Beyond the Lab at TechCrunch Disrupt 2026

The upcoming session, "Building AI Systems When Failure Is Not an Option," features Nathan Michael of Shield AI, Waabi CEO Raquel Urtasun, and Mikell Taylor from General Motors. These experts represent distinct sectors—defense, autonomous trucking, and industrial robotics—each grappling with the challenge of moving beyond the laboratory to ensure reliable operation in unpredictable, high-stakes settings.

Nathan Michael, who oversees Shield AI’s Hivemind autonomy software, brings a background in resilient systems to the discussion, specifically regarding the software's role in U.S. Air Force drone programs. Meanwhile, Raquel Urtasun highlights the necessity of virtual validation, citing the Waabi World simulator as a critical tool for stress-testing autonomous drivers before they reach public roads. Complementing these technical perspectives, Mikell Taylor provides insight into the human-centric design of robotics at General Motors, emphasizing that user trust and workplace integration are as vital as hardware precision.

This panel is part of the broader TechCrunch Disrupt event taking place October 13–15 at San Francisco’s Moscone West. As the industry faces increasing pressure to standardize safety protocols, the conversation aims to define the threshold between experimental technology and mission-ready deployment.