Sensor Fusion Engineer, Machine Learning

Key Skills

Java Programming, Python Programming, Ruby on Rails, Mechanical, Java Programming, Unix Operating system, logo, Offshore,

Job Description

As a low-level Sensor Fusion Engineer at Rivian, you will be part of a team responsible for creating real-time virtual representations of the local environment around the vehicle that includes objects, obstacles, lanes, and drivable space. The team uses raw and processed measurements from sensors on the existing production vehicles in addition to evaluating new sensors and developing algorithms for future vehicle programs. Experience fusing the low/mid level sensor fusion is desired. The ideal candidate would be up to date on the latest advances in machine learning approaches for sensor fusion and multiple object tracking.

Responsibilities

• Implement and evaluate algorithm proof of concepts to support new features
• Work with high-level sensor fusion engineers, to integrate models/algorithms for approved features to the production embedded platform
• Define KPIs and report metrics to verify algorithms meet performance requirements and safety goals
• Root cause and resolve issues
• Support functional safety certification
• Evaluate, and support sourcing efforts for sensors on future vehicle programs
• Follow Agile processes for SW development
• Follow processes to meet ISO26262 and SOTIF requirements
• Work closely with cross-functional stakeholders in systems, controls, simulation, embedded, hardware, and test validation teams

Qualifications

• MS or Ph.D. in Electrical/Mechanical/Aerospace Engineering, Computer Science, Robotics, Mechatronics or a related field
• Proven 2+ years of experiences of algorithm/software development in self-driving or similar industry.
• Good knowledge of radars, LiDARs, cameras, and ultrasonics
• Experience in one or more of the following areas using machine learning/deep learning methods:
• Object detection based on fusion of point cloud sensors and cameras.
• Object detection based on point cloud sensors such as radar and/or LiDAR.
• Depth completion based on point cloud sensor and camera fusion.
• Lane/road detection, segmentation based on fusion of point cloud sensors and cameras.
• Multiple-object-tracking and data association.
• Strong working experience of at least one ML frameworks, such as Tensorflow and PyTorch.
• Strong Python scripting skills.
• Being able to derive reasonable labeling requirements for the ML models, given practical constraints.
• Being able to work in a fast-paced environment and willing to be cross-functional.

Preferred Skills

• Strong C++ programming skills.
• Experience of integrating ML model into real-time production system.
• Knowledge of classical object-tracking and Bayesian framework such as data association, measurement uncertainty, nonlinear filters, etc.
• Knowledge of safety critical systems (ISO 26262)
• Knowledge of verification processes: EV/DV, unit tests, code coverage and static code analysis, SIL/HIL, regression testing, in addition to ML validations.