Adas Interior Perception Systems Engineer

Ford Ford · Auto · Dearborn, MI +1 · Research and Advance Engineering

The role focuses on defining and verifying performance targets for interior perception systems in ADAS, emphasizing statistical methods and consistency across diverse customer populations. It involves developing validation strategies, building application-agnostic frameworks, and collaborating with cross-functional teams and vendors.

What you'd actually do

  1. Define quantitative performance targets. Translate feature-level requirements into verifiable perception subsystem KPIs (e.g., detection and false-positive rates, time-to-detect) and drive alignment on those targets with feature owners and senior leadership.
  2. Characterize system performance across the full range of customer and environmental noise factors (occupant anthropometry, appearance variation, eyewear, lighting, sun load) and develop methodologies to drive consistent feature availability across our customer population.
  3. Design statistically efficient validation strategies (design of experiments, stratified sampling, sequential testing, resampling-based confidence estimation) that characterize system robustness across interacting noise factors while making efficient use of test time and resources.
  4. Build application-agnostic performance and validation frameworks that decouple core perception KPIs from vehicle-specific factors such as camera mounting position, interior geometry, and occupant position, enabling reuse across platforms.
  5. Define real-world usage profiles and exposure models to project bench and track performance into expected field performance for a representative customer population.

Skills

Required

  • Bachelor’s Degree in Electrical Engineering, Computer Science, Mathematics/Statistics, or related field
  • 3+ years of experience decomposing high-level feature requirements into quantitative subsystem KPIs in automotive, aerospace, defense, medical devices, or a similarly safety-relevant domain
  • Demonstrated fluency in applied statistics – hypothesis testing, confidence and tolerance intervals, sample size justification, and reasoning about rare-event rates
  • Experience defining or executing validation strategies under finite test budgets
  • Strong communication skills–particularly the ability to explain and defend a statistical recommendation to non-statistical stakeholders and drive consensus across cross-functional teams

Nice to have

  • Master’s Degree in Electrical Engineering, Computer Science, Mathematics/Statistics, or related field
  • 5+ years of systems engineering experience in a safety-relevant domain
  • Experience with ADAS perception systems and performance evaluation (ROC/DET analysis, operating point selection, confusion matrix decomposition by subgroup)
  • Experience with driver/occupant monitoring or camera-based biometric systems
  • Proficiency in Python (or comparable language) for statistical analysis of large datasets
  • Familiarity with naturalistic driving studies, exposure modeling, or reliability and field-return analysis
  • Understanding of functional safety (ISO 26262) and SOTIF standards and processes
  • Experience with systems modeling and/or requirements management tools (SysML/MagicDraw, Polarion, JAMA, DOORS, etc.) and failure mode analysis techniques
  • Familiarity with requirement formats including INCOSE and EARS
  • A track record of driving consensus on performance targets in environments with competing technical and program pressures
  • Experience presenting and defending performance methodology to executive technical leadership
  • History of building validation frameworks or evaluation methodologies that were adopted beyond their original scope
  • Recognized depth in either applied statistics or perception system design – and working fluency in the other

What the JD emphasized

  • core design objective
  • statistically fluent systems engineer
  • performance targets
  • validation strategies
  • customer population
  • perception subsystem KPIs
  • consistent feature availability
  • statistically efficient validation strategies
  • interacting noise factors
  • application-agnostic performance and validation frameworks
  • real-world usage profiles
  • project bench and track performance into expected field performance
  • cross-functional design reviews
  • statistical characterization of any field issues
  • performance methodology
  • executive technical leadership
  • performance methodology