Vehicle Prognostics - Applied Data Scientist

Ford Ford · Auto · Dearborn, MI +1 · PD Operations and Quality

Applied Data Scientist at Ford focused on developing and deploying prognostic features for vehicle components, using Physics-Informed Machine Learning, C++ edge model deployment, and end-to-end pipeline ownership from simulation to production. The role involves analyzing high-frequency signal processing, multi-sensor fault detection, causal inference, and big data ingestion.

What you'd actually do

  1. Own the process for prognostic feature development from conceptual to feature deployment to our production vehicles.
  2. Pioneer Physics-Informed Machine Learning (PIML): Fuse first-principles physics modeling with advanced machine learning to develop hybrid, high-fidelity prognostic models that capture complex degradation behaviors across both EV and ICE powertrains.
  3. Architect Prognostics & RUL Frameworks: Design and deploy state-of-the-art prognostics models to accurately estimate the Remaining Useful Life (RUL) of critical vehicle subsystems, transforming noisy fleet data into actionable maintenance alerts.
  4. Deploy Edge Models in C++: Translate complex predictive models into highly optimized, low-latency C++ code, bridging the gap between cloud-based data science and resource-constrained on-board vehicle electronic control units (ECUs).
  5. Harness High-Frequency Signal Processing: Architect custom Digital Signal Processing (DSP) pipelines and time-series analytics to extract clean, high-frequency physical signatures from multi-sensor vehicle networks, isolating early-stage wear patterns before they manifest as failures.

Skills

Required

  • Bachelor's in Mechanical, Electrical, Computer Science, Computer engineering, Physics, Mathematics or related fields or a combination of education and equivalent experience
  • 4+ years of experience of practicing statistical methods and their accurate application e.g. ANOVA, principal component analysis, correspondence analysis, k-means clustering, factor analysis, multi-variate analysis, Neural Networks, causal inference, Gaussian regression, etc.
  • 3+ Experience with Python (and related modules), SQL
  • Experience with embedded controls, onboard Diagnostic, Sensor Processing, General First Principles Physics Modeling and simulation using numerical computational tool (e.g. MATLAB, ATI, Simulink)
  • Experience with Digital Signal Processing (DSP) data structures, algorithms, and software engineering principles
  • Self-motivated, strong analytical, excellent interpersonal and communication skills required

Nice to have

  • Master's or PhD in Mechanical, Electrical, Computer Science, Computer engineering, Physics, Mathematics or related fields or a combination of education and equivalent experience
  • Experience in Dynamic Systems, Control, Robotics, Prognostics and Health Management
  • Familiarity working with Automotive prognostics feature development using connected vehicle data.
  • 2+ Experience in application of statistical and machine learning methods e.g., ANOVA, PCA, clustering methods, causal inference, time series forecasting, random forest, multi-variate analysis, neural networks, etc.
  • Expertise in open-source data science technologies such as Python, R, Spark, Hadoop, etc. acquired through college course work, online training and certification or project development.
  • Experience in software development for automotive controls with hands on experience using MATLAB for large scale data and understanding of programming fundamentals and experience with C++ programming in embedded environments. ATI and ETAS calibration tool fa

What the JD emphasized

  • Own the process
  • Deploy Edge Models in C++
  • Own the End-to-End Pipeline (HIL to Production)

Other signals

  • Physics-Informed Machine Learning
  • Remaining Useful Life (RUL) estimation
  • Deploy Edge Models in C++
  • Multi-sensor Fault Detection & Isolation (FDI)
  • Statistical Causal Inference
  • End-to-End Pipeline (HIL to Production)
  • Big Data & Calibration Tools