AD/ADAS Verification and Compliance Hub

Unifies safety-critical requirements management, traceability, and compliance evidence with simulation-driven validation for autonomous driving systems, including LiDAR development, camera optical-effect modeling, and repeatable MIL/SIL/HIL testing.

The Problem

“AD/ADAS compliance verification and virtual validation for safety-critical autonomous driving systems”

Organizations face these key challenges:

1

Fragmented requirements, verification, and change-management workflows across teams and tools

2

Manual creation of traceability matrices and ISO 26262 compliance evidence

3

Limited realism in camera and sensor simulation for optical failure modes

4

High cost and low repeatability of physical validation for AD/ADAS edge cases

Impact When Solved

Reduce manual compliance evidence preparation by 40-70% through automated traceability and report generationIncrease requirement-to-test coverage visibility across LiDAR, camera, and vehicle functionsAccelerate MIL/SIL/HIL validation cycles with orchestrated scenario execution and result triageImprove perception robustness by testing optical and sensor edge cases before road deployment

The Shift

Before AI~85% Manual

Human Does

  • •Collect requirements, hazards, design changes, and test records from separate ALM, simulation, and document workflows
  • •Manually build traceability matrices and assemble ISO 26262 compliance evidence for audits and release reviews
  • •Plan and run LiDAR, camera, and vehicle-level validation across MIL, SIL, and HIL with manual scenario coordination
  • •Review simulation and physical test results to identify gaps, regressions, and perception edge-case failures

Automation

    With AI~75% Automated

    Human Does

    • •Approve requirement interpretations, safety priorities, and release-readiness decisions for AD/ADAS programs
    • •Review and sign off on compliance evidence packages, traceability exceptions, and unresolved validation gaps
    • •Decide how to handle critical perception failures, model limitations, and cross-stage result conflicts

    AI Handles

    • •Continuously link requirements, hazards, tests, simulation artifacts, and change records into a unified traceability view
    • •Detect missing evidence, traceability gaps, impacted items from changes, and likely compliance risks across programs
    • •Orchestrate MIL, SIL, and HIL scenario execution, normalize outputs, and compare results across development stages
    • •Generate audit-ready summaries, traceability reports, and validation evidence drafts for LiDAR, camera, and vehicle functions

    Operating Intelligence

    How it works

    AI surfaces what is hidden in the data.

    Humans do the substantive investigation.

    Closed cases sharpen future detection.

    Confidence83%
    ArchetypeDetect & Investigate
    Shape6-step funnel
    Human gates1
    Autonomy
    67%AI controls 4 of 6 steps

    Who is in control at each step

    Each column marks the operating owner for that step. AI-led actions sit above the divider, human decisions and feedback loops sit below it.

    Loop shapefunnel

    Step 1

    Scan

    Step 2

    Detect

    Step 3

    Assemble Evidence

    Step 4

    Investigate

    Step 5

    Act

    Step 6

    Feedback

    AI lead

    Autonomous execution

    1AI
    2AI
    3AI
    5AI
    gate

    Human lead

    Approval, override, feedback

    4Human
    6↺ Loop
    AI-led step
    Human-controlled step
    Feedback loop
    TL;DR

    AI scans and assembles evidence autonomously. Humans do the substantive investigation. Closed cases improve future scanning.

    The Loop

    6 steps

    1 operating angles mapped

    Operational Depth

    Real-World Use Cases

    ISO 26262-ready ALM for autonomous vehicle LiDAR development

    LeddarTech replaced scattered Word and Excel files with one system that tracks product requirements, tests, and changes for LiDAR used in autonomous vehicles, making safety audits and teamwork much easier.

    Knowledge graph-like traceability and workflow intelligence across requirements, tests, verification, and changesproduction enterprise workflow modernization deployed since 2017; mature process digitization rather than experimental ai.
    10.0

    Simulation-centric virtual validation across MIL/SIL/HIL for automated driving

    Instead of relying mostly on real cars, Stellantis tests automated-driving software in a shared virtual environment that works across different development stages and hardware setups.

    digital twin-style simulation orchestration and cross-platform validationoperational and expanding, with active monitoring of simulation-to-reality correlation.
    10.0

    Camera lens-effect simulation for AD/ADAS perception using Speos plus AVxcelerate

    ZF connects optical simulation with sensor simulation so it can model how a camera lens really behaves, including flaws like flare, before the camera is used in a driving system.

    Optics-informed sensor simulation for perception robustness testingplanned/next-stage capability explicitly described as future enablement within the tool chain.
    9.5

    Free access to this report