Perception & Sensor Fusion Engineer

THE POSITION

Saildrone's vessels use multiple sensors to detect and track objects in the maritime environment. The challenge is turning imperfect observations from different sensors into reliable tracks and useful information for autonomous systems and mission operators.

As a Perception & Sensor Fusion Engineer, you will develop the algorithms and software that combine sensor observations into detections, tracks, and higher-level representations of the environment. Your work may span machine learning, computer vision, tracking, sensor fusion, and active sensing—for example, using information from one sensor to direct a PTZ camera toward an object of interest.

THE RESPONSIBILITIES

• Develop perception and sensor-fusion algorithms that combine observations from multiple sensor modalities.
• Develop and evaluate algorithms for object detection, classification, tracking, and estimation.
• Train, evaluate, and deploy machine-learning models where ML is the right approach to the problem.
• Develop algorithms that use information from one sensor to guide or cue another sensor, including PTZ cameras.
• Analyze real-world sensor data to identify failures, edge cases, and opportunities to improve performance.
• Integrate perception outputs with production software and cloud applications used by mission operators.
• Validate and continuously improve perception capabilities using real operational data.

THE QUALIFICATIONS

• Strong software engineering and algorithm-development skills in Python and/or C++.
• Hands-on experience with perception, computer vision, sensor fusion, tracking, or related robotics algorithms.
• Experience working with real sensor data and developing systems that must perform reliably in uncontrolled environments.
• Strong understanding of one or more areas such as object detection, multi-object tracking, state estimation, sensor fusion, or computer vision.
• Experience developing and evaluating machine-learning models for perception.
• Ability to turn noisy sensor observations into reliable, measurable system outputs.
• Experience taking algorithms from development through integration and validation in a production system.

NICE TO HAVE

• Experience with PyTorch or another modern ML framework.
• Experience with multimodal fusion involving cameras, radar, LiDAR, sonar, or other sensors.
• Experience with active perception, sensor cueing, visual tracking, or camera control.
• Experience deploying perception systems on autonomous vehicles, robotics, aerospace, or other physical systems.