Software Engineering
Ann Arbor, MI, USA
TEAM
An autonomous driving system typically comprises Sensing, Perception, Localization, Planning, and Control components. The World Understanding team owns responsibilities for Sensing, Perception, and Localization within Woven by Toyota for autonomous driving development. Specifically, the Sensor & Calibration sub-team, operating under the World Understanding team, focuses on Intrinsic and Extrinsic calibration of autonomous driving sensors, including Camera, LiDAR, Radar, and IMU. Our scope of work also encompasses the quality management of sensor data for post-processing activities and the operational management of calibration for our autonomous driving test vehicles.
WHO ARE WE LOOKING FOR?
The Sensor & Calibration sub-team is seeking a Software Engineer to support the development of online (dynamic) sensor calibration algorithms for our next-generation autonomous driving system. You will contribute to the design and implementation of online sensor calibration algorithms, help validate calibration quality, and assist in optimizing runtime performance within limited computational resources, ultimately enabling deployment of these solutions to millions of Toyota production vehicles. We require an individual with domain expertise in sensor calibration for autonomous driving systems, a basic understanding of autonomous driving sensor hardware such as Camera, LiDAR, Radar, or IMU, and proven experience developing algorithms and tools for sensor calibration. Furthermore, candidates who exhibit a collaborative, "giver" mindset, consistently seeking opportunities to assist colleagues while maintaining a resolute focus on delivering production-ready solutions, are highly valued.
Contribute to the design and implementation of next-generation online (dynamic) sensor calibration algorithms, addressing deficiencies and helping ensure solutions meet objectives for reliability and quality.
Implement and maintain validation pipelines that verify the online sensor calibration algorithms meet the calibration accuracy requirements specified by Toyota products.
Assist in optimizing sensor calibration algorithms for runtime, accuracy, and robustness, taking into account constraints imposed by limited computational resources.
Collaborate with vehicle engineers and operators to ensure exceptional sensor calibration quality, which is critical for optimal vehicle deployment and the enhancement of autonomous driving performance.
Operate within a high-velocity environment and employ rigorous agile development practices.
Work within a hybrid workspace model, necessitating physical presence in our Ann Arbor (Michigan, US) office three days per week.
A minimum of three years of industrial experience in computer vision or a related field, with demonstrated expertise in topics such as:
Sensor Calibration for IMU, LiDAR, Camera, or Radar
Structure-from-motion
SLAM or Localization
Sensor fusion
3D reconstruction or Multi-view geometry
Proficiency in C++ and Python programming languages.
Practical, hands-on experience in constructing a comprehensive sensor calibration stack for autonomous driving systems.
Profound understanding of autonomous driving sensors, including Camera, LiDAR, Radar, and IMU.
Experience in developing testable and modular code to maintain a robust production environment.
The ability to thrive in a fast-paced environment and collaborate effectively across teams and disciplines, supported by excellent interpersonal skills.
Experience in runtime optimization specifically for Linux and UNIX-like real-time operating systems.
Experience with embedded CUDA development and associated tooling.
A proven track record of deploying Machine Learning (ML) models at scale within autonomous driving or related technological fields.