Principal Applied Scientist Perception, Compass
Core
Define scientific direction for safety-critical perception in autonomous robots, developing predictive models of dynamic environments to replace conservative assumptions with learned risk understandings.
Role type
Principal Applied Scientist (Perception)
Builds
Perception pipelines and predictive models for robot safety systems
Domain
Robotics, Autonomous Systems, Safety-Critical AI
Deliverable
production ML models
Required skills
3D scene understanding, object detection and tracking, motion prediction, occupancy forecasting, semantic scene representation, Python, C++, production-grade code, technical direction, cross-functional influence
Preferred skills
uncertainty quantification, out-of-distribution detection, formal verification, control barrier functions, reachability analysis, real-time perception on edge compute, foundation models, self-supervised learning, functional safety standards, human motion prediction, team leadership
Technologies
LiDAR, depth cameras, RGB, radar, embedded GPUs
Responsibilities
Define long-term scientific vision for safety-critical perception, develop novel perception algorithms for dynamic environments, design perception outputs coupled to safety constraints, research methods to quantify perception uncertainty, architect perception pipelines generalizing across sensor modalities, investigate foundation models for safety-critical tasks, collaborate with controls and planning teams, publish research at top-tier venues, mentor applied scientists and research engineers, influence organizational safety architecture
Seniority
Principal, strategy & mentorship
