Key Notes
LiDAR calibration and cross-sensor alignment expertise are required for this senior individual-contributor role. You will set the 3D sensing architecture while remaining hands-on with algorithms, evaluation and deployment. The full-time, hybrid position can be based in London or Sunnyvale; experience with GTSAM, Ceres Solver or g2o is desirable.
What You'll Work On
- Set 3D sensing architecture and develop LiDAR calibration and registration methods, including geometric and learned approaches to aligning cameras and vehicle sensors.
- Develop temporal alignment and motion compensation, addressing synchronisation errors, rolling-shutter behaviour and LiDAR sweep effects across real vehicle sensor configurations.
- Establish evaluation datasets and metrics, productionise sensing systems and guide engineers through cross-functional integration, deployment and assessment across varied operating conditions.
Why This Role Matters
- Cross-sensor alignment determines the consistency of 3D information used for perception and ground truth, making calibration quality consequential across reconstruction and downstream evaluation workflows.
- Vehicle motion and imperfect sensor timing distort measured geometry, requiring compensation methods that remain reliable across differing sensor configurations and real-world operating conditions.
What They Are Looking For
- Experience: Relevant experience includes LiDAR/camera robotics systems, calibration and registration.
- Technical: Relevant technical capabilities include 3D geometry, state estimation, nonlinear optimisation and timing/motion effects.
- Technical: Relevant technical capabilities include C++ and/or Python with production algorithm deployment and evaluation.
- Preferred: Preferred attributes include SLAM, depth models, GTSAM/Ceres/g2o and dynamic-scene methods.