Build robust localization and mapping systems. Develop and improve SLAM pipelines that enable robots to localize reliably and build accurate maps in real-world factory environments.

Fuse sensor data into reliable state estimates. Work on perception and estimation pipelines that combine cameras, IMUs, and other onboard sensors for robust pose estimation and navigation.

Improve autonomy in challenging environments. Tackle real-world problems such as dynamic lighting, motion blur, fast moving vehicles, and changing environments to make localization and mapping more reliable on hardware.

Develop mapping and optimization systems. Contribute to pose graph optimization, filtering based methods, map maintenance, and dynamic mapping data structures that support long term robot operation.

Prototype fast, deploy faster. Test algorithms directly on live robotic systems, evaluate performance with real sensor data, and iterate quickly based on what actually works on the factory floor.

Debug across the full robotics stack. Troubleshoot software, hardware, and system level issues across sensing, calibration, perception, and navigation pipelines.

Research and apply practical improvements. Explore online sensor calibration strategies, online map optimization techniques, and other methods that improve SLAM robustness in production settings.

Explore and prototype learning based SLAM methods. Investigate data driven approaches that improve robustness across dynamic environments, challenging lighting conditions, fast motion, and other real-world failure cases.

Collaborate across teams. Work closely with robotics, backend, and deployment teams to integrate perception and localization capabilities into larger robotic systems.