Ford Motor Company | Research Engineer – Scientific Machine Learning | Dearborn, MI | United States | BigDataKB.com | 22 Oct 2022

The Core Artificial Intelligence, Machine Learning and Quantum Computing (AI/ML/QC) research group in Ford’s Research and Advanced Engineering organization is leading a process of discovery and democratization of AI-ML across many engineering functions within the Ford product landscape. Our team is seeking highly qualified, motivated candidates to grow our research in scientific machine learning (SciML). In this role you will help integrate novel machine learning approaches such as physics informed neural networks (PINNs) with existing scientific models, often based on differential equations, to accelerate scientific discovery.

As part of the Core AI-ML-QC team you will work with other scientists and engineers to research and build machine learning methods to address scientific challenges in materials discovery, electrochemistry, computational fluid dynamics, heat transfer, electromagnetics, and many more. For these domains, you will be responsible for advancing the state-of-the-art in the general concepts and underlying mathematical constructs of SciML, specifically PINNs. This role requires a strong mathematical background as well as an intimate knowledge of machine learning, and multi-physics problems and their solutions. You will be responsible for identifying high impact problem statements and defining new directions of research. You will also work with practicing engineers to transfer these results into practical applications that help create product and process differentiation for Ford.

The Minimum requirements we seek

• Master’s degree in Electrical Engineering , Computer Engineering or related field
• 1+ years of research experience in engineering (mechanical/chemical/electrical/civil/material), mathematics, or physics that is graduate level research must include numerical solution methods for ODE’s, PDE’s, and related stochastic problem classes.
• 2+ years of experience (may include graduate research) in scientific machine learning techniques and development of surrogates of physical models using physics-informed neural networks (PINNs).
• 1+ years of experience (may include graduate research) investigating new methods for neural-surrogates of PDE’s such as Neural-PDE’s, Fourier Neural Operators, or Physics Informed Neural Operators.
• 2+ years of experience in Python, and experience programming with common AI/ML libraries, such as scikit-learn, TensorFlow or Pytorch and/or neural network architectures such as CNNs, RNNs, autoencoders, or generative models. Familiarity with Julia, JAX, and/or Flux.jl is a plus.

Our Preferred Qualifications

• PHD in Electrical Engineering , Computer Engineering or related field
• Direct experience in the use of PINNs in one or more problem domains such as 2D lithium battery electrochemistry models, CFD, heat transfer, electromagnetics etc.
• Direct experience in the use of scientific machine learning environments like Julia Sim, DeepXDE, NeuralPDE.jl for Julia, and NVIDIA Modulus.
• Demonstrated prior work in the development of transfer learning methods for accelerated physics-informed neural network training.
• Experience with data assimilation methods or other data driven reduced order modeling techniques.
• Experience in leveraging high performance computing, multithreading/parallel computing, and GPUs.
• Knowledge of uncertainty quantification techniques.
• Experience with efficient sampling methods.
• Experience with efficient meshing methods.
• Excellent written/oral communication skills.
• Proven ability to work well with others as part of a diverse global team.

What you will receive in return

As part of the Ford family, you will enjoy excellent compensation and a comprehensive benefits package that includes generous PTO, retirement, savings and stock investment plans, incentive compensation and much more. You will also experience exciting opportunities for professional and personal growth and recognition.

If you have what it takes to help us redefine the future of mobility, we would love to have you join us.

Candidates for positions with Ford Motor Company must be legally authorized to work in the United States on a permanent basis. Verification of employment eligibility will be required at the time of hire.

Visa sponsorship is available for this position.

We are an Equal Opportunity Employer committed to a culturally diverse workforce. All qualified applicants will receive consideration for employment without regard to race, religion, color, age, sex, national origin, sexual orientation, gender identity, disability status or protected veteran status.

For information on Ford’s benefits and compensation, click here: https://corporate.ford.com/content/dam/corporate/us/en-us/documents/careers/2022-benefits-and-comp-GSR-sal-plan-2.pdf

At Ford, the health and safety of our employees is our top priority. Vaccination has been proven to play a critical role in combating COVID-19. As a result, Ford has made the decision to require U.S. salaried employees to be fully vaccinated against COVID-19, unless employees require an accommodation for religious or medical reasons. Being fully vaccinated means that an individual is at least two weeks past their final dose of an authorized COVID-19 vaccine regimen. As a condition of employment, newly hired employees will be required to provide proof of their COVID-19 vaccination or an approved medical or religious exemption

What you will be able to do:

• Be Ford’s lead expert and point of view owner on all topics related to scientific machine learning (SciML) and physics informed neural networks (PINNs).
• Progress state-of-the-art research in SciML and PINNs and their variants in alignment with high priority Ford use cases.
• Research, develop and integrate the latest techniques of SciML into the Ford product landscape.
• Create pretrained surrogates of physical models using PINNs.
• Investigate new methods for neural surrogates of partial differential equations like Fourier Neural Operators.
• Develop and demonstrate transfer learning methods for accelerated physics-informed neural network training.
• Develop generalizable (non-local) SciML solutions for providing high fidelity predictions over the full performance domain of interest.
• Maintain, grow, and leverage relationships with leading universities and research bodies on the topic of SciML.
• Present and publish in reputed conferences and high impact peer reviewed Journals.
• Publish internally and disseminate knowledge within the company through teaching, seminars, and internal learning and development programs.
• Develop research portfolio, for internal research as well as for research with leading universities, to allow Ford to maintain a leadership position in the industry.