Simulation is central to ensuring the safety and scalability of autonomous vehicles at Zoox. In this role, you will develop metrics and evaluation pipelines to assess simulator fidelity and guide improvements to our GenAI-powered simulator of virtual worlds with dynamic agents and realistic sensor data.
You’ll collaborate with autonomy, safety, and simulation teams to define quantitative validation criteria for simulator fidelity, and use exploratory data analysis to root-cause gaps in simulator performance. Your work will directly support Zoox scaling autonomous driving across multiple cities in increasingly complex environments.
In this role, you will:
Define and implement metrics to measure the accuracy of simulated sensors, agents, and virtual environments.
Drive consensus on validation methodologies with cross-functional teams in Autonomy, Safety, and Simulation.
Develop robust evaluation strategies for simulator fidelity, mine datasets, and lead their implementation in large-scale evaluation pipelines.
Write design proposals and drive execution while proactively communicating risks to partner teams.
Qualifications:
5+ years of experience in quantitative evaluation, autonomous robotics, or sensor data analysis (lidar, camera, radar)
Proficiency in Python and data analysis (e.g., Pandas, NumPy, SQL), with a strong foundation in data visualization
Working knowledge of applied statistical methods (e.g., histograms, probability distributions, simple hypothesis testing)
Strong communication skills and thrives in dynamic environments by proactively seeking clarity
BS, MS, or Ph.D. in Computer Science, Robotics, or a related field
Bonus Qualifications:
Experience with autonomous vehicle software stack (e.g., Perception, Prediction, Planning)
Experience implementing large-scale evaluation pipelines
Experience using applied statistics for system validation (e.g., confidence intervals, p-values)
