The Machine Learning Motion Planning team specializes in the fusion of Machine Learning and Classical Methods, researches and develops new heuristic and deep learning approaches, and builds deep neural networks, algorithms, and software to make an impact across navigation, behavior, route planning, as well as trajectory optimization, numerical optimization, and model predictive control. The team owns the full development cycle of ML Planner, from ideation, experimentation, validation, simulation, in-vehicle testing, and final deployment into production.
Motional's Machine Learning team has produced groundbreaking advancements in the autonomous vehicle industry, including nuScenes (https://www.nuscenes.org), PointPillars (https://arxiv.org/abs/1812.05784), and PointPainting (http://arxiv.org/abs/1911.10150)
What You'll Do:
- Contribute to cutting-edge motion planning systems with Machine Learning and optimization-based methods.
- Develop a variety of algorithms for planning for autonomous driving, including but not limited to search-based methods, sampling-based methods, optimization-based methods, linear-temporal logic, and decision-making under uncertainty.
- Develop core deep learning algorithms for efficient training and testing pipelines.
- Use your top-notch software development expertise to inspire others to develop better software practices and principles.
- Design and build a robust and scalable codebase that enables rapid exploration and evaluation of different motion planning approaches and algorithms.
- Interface with perception and prediction components upstream and trajectory optimization and tracking & control components downstream
- Refine and improve the data-driven model according to the performance of AV in real-world complex environment
- Deploy solutions directly to an autonomous vehicle and analyze their impact
- Provide a vision for the team and our system—guide and mentor junior team members to develop a culture of product-focused engineering, research, and development.
What You'll Bring:
- Bachelor's, Masters, or PhD degree preferred in Machine Learning, Robotics, Computer Science, Computer Engineering, Mechanical Engineering, Applied Mathematics, Statistics, or a related field
- 5+ years of software development experience
- 2+ years of C++ software development experience preferred
- Experience designing, training, analyzing, and deploying neural networks with PyTorch or other deep learning frameworks
- Experience owning and leading technical development on features from problem formulation through implementation and deployment
- Thirst for knowledge and continuous innovation
