Perception Software Engineer

We are seeking a hands-on Perception Software Engineer to join the Perception team. In this role, you will use your expertise to shape the perception stack as we scale, incorporating and developing state of the art algorithms and techniques. You’ll leverage advanced Artificial Intelligence (AI) and Machine Learning (ML) models and techniques to deploy and scale safe autonomous perception systems.

• Design, develop, and optimize perception and machine learning (ML) systems for autonomous trucking applications.
• Implement and evaluate real-time perception algorithms for 3D object detection, tracking, and scene understanding using multimodal sensor data (LiDAR, camera, and radar).
• Develop new deep learning architectures to enhance perception model accuracy and reliability under diverse driving conditions.
• Integrate perception models into real-time autonomous driving stacks and deploy models to production systems running on autonomous trucks.
• Lead development of offboard and onboard perception pipelines, including association evaluators, auto labeling frameworks, and ground estimation networks.
• Design and implement large-scale evaluation tools for tracking and association, improving error diagnosis and reducing false positive braking events.
• Conduct model optimization for real-time inference using TensorRT and ONNX and implement efficient multi-threaded data processing in C++ for perception modules.
• Collaborate cross-functionally with planning, safety, and simulation teams to improve pedestrian safety cases and vehicle behavior around vulnerable road users.
• Design and execute experiments to analyze perception failures, build evaluation suites, and refine MLdriven improvements to perception performance.

• Master of Science degree in Robotic Systems Development, Robotics Engineering, Mechatronics Engineering, Computer Science or a closely related field.
• At least ONE (1) year of experience in the job offered or at least ONE (1) year of experience in the following:

1.  Computer Vision & Deep Learning
2.  Designing and implementing image segmentation ML architectures
3.  Image preprocessing and augmentation
4.  Python and ML libraries, including TensorFlow, Keras, OpenCV, Scikit-image, NumPy, and SciPy
5.  Train and optimize ML models, including hyperparameter tuning, dropout scheduling, loss weighting, and evaluation metrics
6.  TensorBoard