NREL HPC enabled deep reinforcement learning for autonomous vehicle control in Golden, Colorado
HPC enabled deep reinforcement learning for autonomous vehicle control
CO - Golden
Postdoc (Fixed Term)
Hours Per Week
The Complex Systems Simulation and Optimization Group in the NREL Computational Science Center has an opening for a full-time postdoctoral researcher in HPC enabled deep reinforcement learning (DRL) for autonomous vehicle (AV) control. The researcher will address the broad spectrum of challenges involved in improving AV control via large-scale HPC, including but not limited to perception (e.g., object detection and trajectory estimation from lidar and visual fields), controls (the final decisions sent to the gears and motors) and connectivity (e.g., optimal energy usage and traffic mitigation by coordinated movement). This project is a collaboration between multiple National Laboratories and Industry partners and has immediate relevance for widespread deployment of AVs.
We are looking for a dynamic, motivated researcher with a strong technical background and an interest in the mission of NREL. The successful candidate will collaborate with NREL staff and researchers, other national labs, universities, and industry partners on efforts to develop and apply AI at scale to connected and autonomous vehicles. In particular, the candidate will:
Identify characteristics and representative data set that enables off line deep reinforcement learning based vehicle control.
Identify and develop neural network architecture components and characteristics required for use in HPC-enabled deep reinforcement learning based vehicle control.
Identify and develop appropriate learning algorithms for HPC-enabled deep reinforcement learning based vehicle control.
Understand state-of-the-art AV control design.
Collaborate with NREL researchers to design and implement deep learning algorithms and computational experiments for AV control applications, including the implementation and execution of workflows on HPC systems.
Develop theories and algorithms to appropriately incorporate DRL into state-of-the-art AV control.
Conduct research in physics-informed deep learning for optimal hybridization of deep learning and model-based methods.
Evaluate and track state-of-the-art of deep learning algorithms, especially deep reinforcement learning, and their implementations and use on advanced and emerging computing architectures.
Author, present, and assist in the preparation of technical papers, reports and conference proceedings on topics related to deep reinforcement learning applied to autonomous vehicle control.
Required Education, Experience, and Skills
Must be a recent PhD graduate within the last three years.
Expertise in deep learning (neural networks, including knowledge of core mathematical underpinnings), reinforcement learning (esp., for continuous and/or large state-action spaces).
Expertise in state-of-the-art autonomous vehicle control algorithms.
Experience with classical control theory, model predictive control, dynamic programming.
Strong background in mathematics, statistics, and probability.
Strong background in physics and engineering.
Strong programming skills, especially experience using deep learning frameworks (e.g. TensorFlow, PyTorch, dopamine, etc) on high performance computing/GPU platforms.
Please note that in order to be considered an applicant for any position at NREL you must submit an application form for each position for which you believe you are qualified. Applications are not kept on file for future positions. Please include a cover letter and resume with each position application.
NREL is dedicated to the principles of equal employment opportunity. NREL promotes a work environment that does not discriminate against workers or job applicants and prohibits unlawful discrimination on the basis of race, color, religion, sex, national origin, disability, age, marital status, ancestry, actual or perceived sexual orientation, or veteran status, including special disabled veterans.
NREL validates right to work using E-Verify. NREL will provide the Social Security Administration (SSA) and, if necessary, the Department of Homeland Security (DHS), with information from each new employee’s Form I-9 to confirm work authorization. For additional information, click here at http://www.nrel.gov/careers/employment-policies.html .
The National Renewable Energy Laboratory (NREL) is a leader in the U.S. Department of Energy’s effort to secure an environmentally and economically sustainable energy future. With locations in Golden and Boulder, Colorado, and a satellite office in Washington, D.C., NREL is the primary laboratory for research, development, and deployment of renewable energy technologies in the United States.
Please review the information on our Hiring Process at https://www.nrel.gov/careers/hiring-process.html website before you create an account and apply for a job. We also hope you will learn more about NREL at https://www.nrel.gov/about/ , visit our Careers site at https://www.nrel.gov/careers/ , and continue to search for job opportunities at https://nrel.wd5.myworkdayjobs.com/NREL at the lab.