From simulation as a sandbox to simulation as a proving ground, advances in simulation realism and scene reconstruction are fundamentally changing how autonomous systems are tested and validated at scale. Developers can now transform drive logs into high-fidelity, simulation-capable digital twins, with full control to run open- and closed-loop testing scenarios. This shift enables teams to scale validation in simulation where they previously relied on costly and time-consuming real-world testing. The result: faster iteration cycles, reduced development costs and greater confidence in system performance. This talk explores the technologies enabling this transformation and their real-world impact.

• How advances in simulation realism are driving a shift in AV testing and validation workflows
• What it takes to generate simulation-ready digital twins from drive logs
• Why simulation-first validation reduces cost, accelerates iteration and builds trust
• Lessons from teams adopting simulation as a core validation tool