Functional Safety AI Principal Engineer

In this position...

Connect AI functional safety to software and quality engineering:

Implement guardrails for safety within existing quality requirements enforced through the same development pipeline, quality gates, and metrics.

Drive system safety architecture and assurance strategy for AI across Ford's ADAS/AD, driver-monitoring, and connected/embodied-AI programs;

Lead adoption of emerging AI-safety standards and process frameworks and prepares Ford for AI-specific regulation;

Principal technical authority bridging architecture, safety, software quality, AI/ML engineering, and program execution.

Day-to-day technical leadership on active programs, forward-looking work that matures Ford's AI-safety, security and quality methods, tooling, and standards posture.

Assess AI maturity within projects, process frameworks and Ford infrastructure to develop an AI in FuSa Health Dashboard.

Work with Ford Connected Vehicle Software and Product Data Integrity, and supplier engineering teams to identify safety related process and work product gaps and recommend best practice solutions.

Develop a capability maturity plan working with engineering stakeholders.

Report to the Director of the Office of Functional Safety Assurance (OFSA).

Define and deploy Ford's strategy for AI in safety critical systems, ensuring that AI- and ML-enabled vehicle features and systems are demonstrably safe, certifiable, and deployable

What you'll do:

• FuSa/AI Technical Leadership: Apply, demonstrate and champion best in class methods, tools, practices to ensure embedded software achieves safety.
• AI in FuSa Software Health Dashboard Ownership: Lead the development of the Health Dashboard for the enterprise, including its metrics and criteria for maturity. Develop AI enhanced tracking and analytic tools to determine status and priority actions.
• Strategic leadership: Provide strategic direction and leadership to the safety community, fostering a culture of collaboration, innovation, and excellence. Design and drive safe AI enablement strategies into cross-organizational teams.
• Ownership of the AI in FuSa strategy: Responsible for the definition and deployment of maturity model, framework for safe AI development and AI enhanced confirmation.
• Cross-functional collaboration: Collaborate closely with cross-functional teams (including product/program management, validation and verification, software, hardware, design, integration, quality, and architecture) to drive execution to program timelines. Cross-functional collaboration: Collaborate closely with cross-functional teams including product engineering, program management, validation and verification, software, hardware, design, integration, quality, and architecture to drive maturity of enterprise AI/FuSa Health metrics.
• FuSa Software related escapes: Apply system engineering, risk analysis, trouble shooting and root cause analysis methods to find software related issues and close the loop to process, methods and tools gaps.
• Alignment of FuSa, Cyber and Quality Assurance: Align and exploit synergies with SQA where this helps drive FuSa maturity.
• Continuous improvement: Drive continuous improvement initiatives to optimize software development and deployment processes, tools, and methodologies, enhancing effectiveness.
• Communication: Provide consistent health status, insights, risks and opportunities to executive leadership and stakeholders.
• Safety Strategy, Roadmap & Governance. Translate Ford's AI-safety vision into an executable multi-year roadmap across ADAS/AD, driver monitoring, and connected/embodied-AI features; align near-, mid-, and long-term horizons and ensure clean transitions between them. Establish the AI functional-safety governance model: decision gates, safety sign-off authority, escalation paths, and accountability across programs and functions. Define and steward organization-wide policies, guidelines, and reference workflows for safe AI/ML development. Provide structured feedback to program and platform leadership on safety posture, gaps, and where investment is needed. Serve as design authority / final technical safety sign-off for AI safety concepts on assigned programs.
• AI/ML System Safety Architecture & Concept. Own the system safety architecture for AI-driven features: define the safety concept, allocate safety requirements, and specify fail-safe / fail-operational behavior, degradation strategies, and minimal-risk-condition (MRC) attainment. Translate learned/AI behavior into verifiable safety requirements, constraints, and runtime safety mechanisms (safety envelopes, monitors, doer/guardian architectures). Define architectural patterns for redundancy, diversity, plausibility checking, and out-of-distribution detection at vehicle and subsystem level. Specify the interaction between AI components and classical safety mechanisms (safety monitors, fallback controllers, arbitration). Establish safety-driven requirements across the data, model, and toolchain lifecycle.
• Safety of the Intended Function (SOTIF) & Performance Limitations. Lead SOTIF (ISO 21448) activities for AI-enabled features: triggering-condition analysis, scenario coverage, and residual-risk evaluation. Define ODD-based safety requirements and the evidence needed to demonstrate acceptable performance within the operational design domain and safe behavior at its edges. Drive systematic identification and mitigation of AI-specific hazards: distributional shift, edge cases, specification insufficiency, and emergent behavior.
• AI Safety Analysis, Verification & Validation. Apply and mature model validation, robustness and adversarial testing, fault injection, and runtime monitoring; identify gaps and feed them back into design. Lead hazard and safety analyses tailored to AI (STPA, HARA, FMEA/FMEDA) and integrate them with system-level analyses. Define validation strategies and acceptance criteria: statistical, scenario-based, simulation, and sim-to-real evidence validity. Establish metrics and KPIs for AI safety performance and assurance completeness.
• Software & AI/ML Quality Engineering. Embed AI safety requirements and acceptance criteria into the software quality management system and stage-gate quality reviews, so safety is enforced through the quality pipeline rather than maintained alongside it. Extend software quality practice to ML artifacts: treat data, models, and pipelines as controlled quality items with versioning, reproducibility, traceability, and defined quality gates (MLOps quality). Define and track quality metrics for AI-enabled software: code quality and coverage, model performance and robustness scores, data-quality indicators, and defect/escape rates; drive them into release criteria and quality dashboards. Align AI/ML development with Automotive SPICE (ASPICE) and its emerging ML/AI extensions; help mature process capability for data management, model training, and model validation. Support and interface with quality and process assessments and audits (ASPICE, functional-safety audits, tool confidence / tool qualification), and drive findings to closure with corrective actions. Establish field-quality feedback loops (production monitoring, drift detection, and incident analysis) that feed back into both quality and safety improvement.
• Assurance Cases & Safety Argumentation. Develop and maintain structured assurance/safety cases (e.g., GSN) for AI-enabled systems, demonstrating absence of unreasonable risk. Integrate ISO 26262, ISO 21448, ISO/PAS 8800, and UL 4600 evidence into a coherent, auditable argument. Prepare safety documentation and readiness for internal reviews, independent assessors, certification bodies, and regulators.
• Standards, Process & Software-Quality Integration. Lead adoption of the emerging AI-safety standards portfolio into Ford's product-development and safety lifecycle: ISO/PAS 8800 (safety and AI for road vehicles), ISO/IEC TR 5469 (functional safety and AI systems), ISO/IEC 23894 (AI risk management), and UL 4600 (autonomous-product safety cases). Anchor process and product quality to the software-quality standards base: Automotive SPICE (and its ML/AI extensions), ISO/IEC/IEEE 12207 (Software life cycle processes), ISO/IEC 25059 (AI-system quality model), and ISO/IEC 5338 (AI-system lifecycle), together with ISO 26262 Parts 6 and 8 for software development and supporting processes. Integrate these with the established automotive safety and cybersecurity baseline: ISO 26262, ISO 21448 (SOTIF), ISO/SAE 21434, and the ISO/TS 5083 (AV Safety) framework for automated driving. Draw on adjacent AI-governance and risk frameworks where useful, such as the NIST AI Risk Management Framework (its Govern / Map / Measure / Manage functions), ISO/IEC 42001 (AI management systems), and ISO/IEC 5259 (data quality for ML). Define which standard governs which claim in the safety and quality argument, and update development processes, work products, and tailoring guidance to incorporate AI-specific requirements. Represent Ford in external standards development (ISO, SAE, UNECE) and translate emerging requirements into internal readiness.
• Regulatory Readiness & Compliance Strategy. Prepare Ford to meet AI-specific regulation, with the EU AI Act (Regulation (EU) 2024/1689) as the anchor: interpret how its high-risk obligations (risk management, data governance, technical documentation, human oversight, robustness, logging, and transparency) apply to Ford's AI-enabled systems. Track how those obligations reach automotive AI through the vehicle type-approval route (Annex I / Regulation (EU) 2018/858 and the General Safety Regulation (EU) 2019/2144). Maintain a forward view of UNECE activity (UN Regs No. 157 and 152, the DCAS work, and the automated-driving framework) and of US frameworks (NHTSA guidance, FMVSS), and advise programs on divergence across markets. Structure Ford's AI risk-governance program around a recognized backbone such as the NIST AI Risk Management Framework, using its crosswalks to the EU AI Act and ISO/IEC 23894 (guidance on risk management) so governance activities (govern, map, measure, manage) stay consistent across markets and the resulting records are reusable as regulatory evidence. Establish the compliance evidence and documentation trail (technical files, risk-management records, data-governance artifacts, and post-market monitoring) that type-approval authorities and market surveillance will expect for AI features, aligned to the assurance case so evidence is produced once and reused. Advise leadership on regulatory strategy, timelines, and readiness gaps; serve as internal point of contact for AI-regulatory questions and, where appropriate, as external liaison.
• Cross-Functional & Supplier / Tier-1 Collaboration. Bridge architecture, safety, software quality and SQA, test engineering, AI/ML engineering, systems, and program-execution teams so safety and quality are designed in rather than bolted on. Define and flow down AI safety requirements to suppliers and Tier-1 partners; review supplier safety cases and evidence, and own the relevant interface/DIA agreements, including the provider/deployer responsibility split that EU AI-Act conformity turns on. Coordinate AI safety across product-development milestones, applications, and platform programs. Support partner-, customer-, and assessor-facing safety discussions.
• Technical Leadership, Mentorship & Capability Building. Serve as principal technical authority and mentor for AI functional safety; grow internal competency across the organization. Build and curate the methods, tools, templates, and training that scale AI-safety practice. Lead and grow a dedicated AI-safety team: set direction, ensure delivery to full potential, and create clear career paths. (optional; include only if the role carries people-management scope). Represent Ford's AI-safety thought leadership internally and externally through publications, conferences, and working groups. (optional).
• Innovation & Applied Research. Drive innovation projects for safe AI aligned to the roadmap; shepherd concepts from research through productization. Evaluate emerging AI-safety methods, tools, and academic advances for applicability to Ford programs. Pilot novel assurance techniques (runtime assurance, formal methods,

What you'll do:

• FuSa/AI Technical Leadership: Apply, demonstrate and champion best in class methods, tools, practices to ensure embedded software achieves safety.
• AI in FuSa Software Health Dashboard Ownership: Lead the development of the Health Dashboard for the enterprise, including its metrics and criteria for maturity. Develop AI enhanced tracking and analytic tools to determine status and priority actions.
• Strategic leadership: Provide strategic direction and leadership to the safety community, fostering a culture of collaboration, innovation, and excellence. Design and drive safe AI enablement strategies into cross-organizational teams.
• Ownership of the AI in FuSa strategy: Responsible for the definition and deployment of maturity model, framework for safe AI development and AI enhanced confirmation.
• Cross-functional collaboration: Collaborate closely with cross-functional teams (including product/program management, validation and verification, software, hardware, design, integration, quality, and architecture) to drive execution to program timelines. Cross-functional collaboration: Collaborate closely with cross-functional teams including product engineering, program management, validation and verification, software, hardware, design, integration, quality, and architecture to drive maturity of enterprise AI/FuSa Health metrics.
• FuSa Software related escapes: Apply system engineering, risk analysis, trouble shooting and root cause analysis methods to find software related issues and close the loop to process, methods and tools gaps.
• Alignment of FuSa, Cyber and Quality Assurance: Align and exploit synergies with SQA where this helps drive FuSa maturity.
• Continuous improvement: Drive continuous improvement initiatives to optimize software development and deployment processes, tools, and methodologies, enhancing effectiveness.
• Communication: Provide consistent health status, insights, risks and opportunities to executive leadership and stakeholders.
• Safety Strategy, Roadmap & Governance. Translate Ford's AI-safety vision into an executable multi-year roadmap across ADAS/AD, driver monitoring, and connected/embodied-AI features; align near-, mid-, and long-term horizons and ensure clean transitions between them. Establish the AI functional-safety governance model: decision gates, safety sign-off authority, escalation paths, and accountability across programs and functions. Define and steward organization-wide policies, guidelines, and reference workflows for safe AI/ML development. Provide structured feedback to program and platform leadership on safety posture, gaps, and where investment is needed. Serve as design authority / final technical safety sign-off for AI safety concepts on assigned programs.
• AI/ML System Safety Architecture & Concept. Own the system safety architecture for AI-driven features: define the safety concept, allocate safety requirements, and specify fail-safe / fail-operational behavior, degradation strategies, and minimal-risk-condition (MRC) attainment. Translate learned/AI behavior into verifiable safety requirements, constraints, and runtime safety mechanisms (safety envelopes, monitors, doer/guardian architectures). Define architectural patterns for redundancy, diversity, plausibility checking, and out-of-distribution detection at vehicle and subsystem level. Specify the interaction between AI components and classical safety mechanisms (safety monitors, fallback controllers, arbitration). Establish safety-driven requirements across the data, model, and toolchain lifecycle.
• Safety of the Intended Function (SOTIF) & Performance Limitations. Lead SOTIF (ISO 21448) activities for AI-enabled features: triggering-condition analysis, scenario coverage, and residual-risk evaluation. Define ODD-based safety requirements and the evidence needed to demonstrate acceptable performance within the operational design domain and safe behavior at its edges. Drive systematic identification and mitigation of AI-specific hazards: distributional shift, edge cases, specification insufficiency, and emergent behavior.
• AI Safety Analysis, Verification & Validation. Apply and mature model validation, robustness and adversarial testing, fault injection, and runtime monitoring; identify gaps and feed them back into design. Lead hazard and safety analyses tailored to AI (STPA, HARA, FMEA/FMEDA) and integrate them with system-level analyses. Define validation strategies and acceptance criteria: statistical, scenario-based, simulation, and sim-to-real evidence validity. Establish metrics and KPIs for AI safety performance and assurance completeness.
• Software & AI/ML Quality Engineering. Embed AI safety requirements and acceptance criteria into the software quality management system and stage-gate quality reviews, so safety is enforced through the quality pipeline rather than maintained alongside it. Extend software quality practice to ML artifacts: treat data, models, and pipelines as controlled quality items with versioning, reproducibility, traceability, and defined quality gates (MLOps quality). Define and track quality metrics for AI-enabled software: code quality and coverage, model performance and robustness scores, data-quality indicators, and defect/escape rates; drive them into release criteria and quality dashboards. Align AI/ML development with Automotive SPICE (ASPICE) and its emerging ML/AI extensions; help mature process capability for data management, model training, and model validation. Support and interface with quality and process assessments and audits (ASPICE, functional-safety audits, tool confidence / tool qualification), and drive findings to closure with corrective actions. Establish field-quality feedback loops (production monitoring, drift detection, and incident analysis) that feed back into both quality and safety improvement.
• Assurance Cases & Safety Argumentation. Develop and maintain structured assurance/safety cases (e.g., GSN) for AI-enabled systems, demonstrating absence of unreasonable risk. Integrate ISO 26262, ISO 21448, ISO/PAS 8800, and UL 4600 evidence into a coherent, auditable argument. Prepare safety documentation and readiness for internal reviews, independent assessors, certification bodies, and regulators.
• Standards, Process & Software-Quality Integration. Lead adoption of the emerging AI-safety standards portfolio into Ford's product-development and safety lifecycle: ISO/PAS 8800 (safety and AI for road vehicles), ISO/IEC TR 5469 (functional safety and AI systems), ISO/IEC 23894 (AI risk management), and UL 4600 (autonomous-product safety cases). Anchor process and product quality to the software-quality standards base: Automotive SPICE (and its ML/AI extensions), ISO/IEC/IEEE 12207 (Software life cycle processes), ISO/IEC 25059 (AI-system quality model), and ISO/IEC 5338 (AI-system lifecycle), together with ISO 26262 Parts 6 and 8 for software development and supporting processes. Integrate these with the established automotive safety and cybersecurity baseline: ISO 26262, ISO 21448 (SOTIF), ISO/SAE 21434, and the ISO/TS 5083 (AV Safety) framework for automated driving. Draw on adjacent AI-governance and risk frameworks where useful, such as the NIST AI Risk Management Framework (its Govern / Map / Measure / Manage functions), ISO/IEC 42001 (AI management systems), and ISO/IEC 5259 (data quality for ML). Define which standard governs which claim in the safety and quality argument, and update development processes, work products, and tailoring guidance to incorporate AI-specific requirements. Represent Ford in external standards development (ISO, SAE, UNECE) and translate emerging requirements into internal readiness.
• Regulatory Readiness & Compliance Strategy. Prepare Ford to meet AI-specific regulation, with the EU AI Act (Regulation (EU) 2024/1689) as the anchor: interpret how its high-risk obligations (risk management, data governance, technical documentation, human oversight, robustness, logging, and transparency) apply to Ford's AI-enabled systems. Track how those obligations reach automotive AI through the vehicle type-approval route (Annex I / Regulation (EU) 2018/858 and the General Safety Regulation (EU) 2019/2144). Maintain a forward view of UNECE activity (UN Regs No. 157 and 152, the DCAS work, and the automated-driving framework) and of US frameworks (NHTSA guidance, FMVSS), and advise programs on divergence across markets. Structure Ford's AI risk-governance program around a recognized backbone such as the NIST AI Risk Management Framework, using its crosswalks to the EU AI Act and ISO/IEC 23894 (guidance on risk management) so governance activities (govern, map, measure, manage) stay consistent across markets and the resulting records are reusable as regulatory evidence. Establish the compliance evidence and documentation trail (technical files, risk-management records, data-governance artifacts, and post-market monitoring) that type-approval authorities and market surveillance will expect for AI features, aligned to the assurance case so evidence is produced once and reused. Advise leadership on regulatory strategy, timelines, and readiness gaps; serve as internal point of contact for AI-regulatory questions and, where appropriate, as external liaison.
• Cross-Functional & Supplier / Tier-1 Collaboration. Bridge architecture, safety, software quality and SQA, test engineering, AI/ML engineering, systems, and program-execution teams so safety and quality are designed in rather than bolted on. Define and flow down AI safety requirements to suppliers and Tier-1 partners; review supplier safety cases and evidence, and own the relevant interface/DIA agreements, including the provider/deployer responsibility split that EU AI-Act conformity turns on. Coordinate AI safety across product-development milestones, applications, and platform programs. Support partner-, customer-, and assessor-facing safety discussions.
• Technical Leadership, Mentorship & Capability Building. Serve as principal technical authority and mentor for AI functional safety; grow internal competency across the organization. Build and curate the methods, tools, templates, and training that scale AI-safety practice. Lead and grow a dedicated AI-safety team: set direction, ensure delivery to full potential, and create clear career paths. (optional; include only if the role carries people-management scope). Represent Ford's AI-safety thought leadership internally and externally through publications, conferences, and working groups. (optional).
• Innovation & Applied Research. Drive innovation projects for safe AI aligned to the roadmap; shepherd concepts from research through productization. Evaluate emerging AI-safety methods, tools, and academic advances for applicability to Ford programs. Pilot novel assurance techniques (runtime assurance, formal methods,