Part 2: Virtual ADAS Testing and Software Validation for AEB Systems

Vandan Babu O M S S N, Sujith Kudupudi
06:00 MINS
October 24, 2025

Autonomous Emergency Braking (AEB) is one of the most crucial technologies shaping the future of safe and intelligent mobility. While functional safety lays the foundation for building reliable braking systems, bringing that intelligence to life requires more than just design—it demands precise software development and extensive virtual validation.

This article is a continuation of Part 1, where we explored the fundamentals of functional safety in AEB systems, ISO 26262 compliance, and how tools like Ansys Medini Analyze help engineers design safer architectures through HARA, FTA, and FMEA analyses.

If you haven’t read Part 1 yet, we recommend starting there—it provides the essential context on how AEB systems are built to detect risks, assess hazards, and maintain safety under all conditions.

In Part 2, we take the next step: discovering how Ansys SCADE and Ansys AVxcelerate bridge the gap between design and real-world performance. You’ll see how model-based development, certified code generation, and physics-accurate virtual testing are redefining the future of ADAS validation.

The Challenge of Manual Coding for ISO 26262 Compliance

Developing an AEB system by manually coding the control algorithm is a complex and error-prone process. Hand-coded software can quickly become difficult to manage, especially as the system grows more complex with the addition of various sensors, algorithms, and safety checks. Furthermore, ensuring that the code complies with ISO 26262 — which requires exhaustive validation, traceability, and fault tolerance — adds an additional layer of difficulty.

Manual coding often lacks built-in traceability, making it hard to link code to specific requirements or safety standards. Achieving certification for ISO 26262 compliance requires extensive testing, documentation, and analysis, making the manual approach both time-consuming and prone to human error. This is where model-based development tools like ANSYS SCADE come into play, providing a more efficient and reliable way to meet safety standards.

ANSYS SCADE and the Value It Adds

To implement the AEB control logic, we developed a model in ANSYS SCADE Suite — a tool specifically designed for developing safety-critical embedded software.

SCADE Model Overview

Fig. 1 Top level view of AEB model developed in Ansys SCADE Suite

Fig2: Simulation of AEB Model – State machine

Fig3: Model based development using Ansys SCADE Suite

Why SCADE is the Right Fit for AEB Development

SCADE brings several strengths that make it especially well-suited for developing AEB and similar ADAS functions:

Model-Based Design with Formal Semantics
Simulation & Early Validation
Traceability to Requirements & Safety Analyses
Certified Code Generation: SCADE’s automatic code generator is qualified up to ISO 26262 ASIL D, allowing you to generate production-ready code directly from the model — reducing both manual coding effort and verification time.

Exporting and Validating the Algorithm

To support broader testing and integration, the SCADE model can be exported in multiple formats:

In our workflow, we wrapped the SCADE model in Python and integrated it into AVXcelerate to validate the AEB logic against synthetic sensor inputs and real-time driving scenarios.

Once the algorithm is validated at the model level, SCADE allows you to generate ISO 26262-certified code that is functionally equivalent to the validated model. This means there’s no need for additional software-in-the-loop (SIL) comparisons or duplicative verification activities on the generated code. The model and the code are one and the same — helping you save time, reduce errors, and streamline certification.

Virtual Test Grounds & why it matters

Ansys AVxcelerate and Ansys SCADE together form a powerful ecosystem that accelerates AEB development from concept to certification. While Ansys Avxcelerate provides physics-based sensor simulation and virtual scenario validation, Ansys SCADE enables ISO 26262-certified model-based software development. Together, they create a continuous workflow that bridges perception, decision, and control — transforming how ADAS systems are designed, tested, and deployed.

Traditionally, AEB validation depended on physical proving grounds, while effective, these tests are expensive, time-consuming, and limited in their ability to cover edge cases. The shift toward digital validation has reshaped how OEM’s approach ADAS development-With virtual testing becoming a cornerstone of modern safety system validation.

Physics-Based Simulation: Why It Matters for ADAS

Many virtual tools rely on ideal sensor outputs – simplified detections without environmental noise, distortion, or real-world variability. While these models are helpful for prototyping, they fail to reveal how algorithms behave in challenging, real-world conditions such as reduced visibility, glare, or sensor noise.

Ansys AVxcelerate sets itself apart with true physics-based simulation, modeling how light, radar waves, LiDAR beams, and thermal radiation interact with materials, weather, and lighting conditions. For AEB development, this realism is essential, as accurate perception directly influences braking decisions and safety outcomes.

With physics-based outputs, AEB testing becomes more than a functional verification step — it becomes a pathway to developing robust, safety-compliant algorithms that can withstand real-world complexities.

Fig: Ground Truth Data from Avxcelerate

From Proving Grounds to Virtual Proving Grounds

Physical providing grounds, while reliable, are limited by repeatability, cost, and safety concerns. Conditions such as weather and lighting cannot be perfectly controlled and replicating rare or dangerous events (like a pedestrian suddenly crossing in front of the vehicle) is risky.

Ansys AVxcelerate transforms this process by enabling engineers to reproduce Euro NCAP test cases — such as AEB City, Pedestrian, and VRU — virtually, while also extending into edge cases that physical testing cannot safely replicate. Every scenario can be repeated under controlled, physics-based conditions for consistent, traceable validation.

AEB Testing Workflow with Ansys AVxcelerate

The AEB validation workflow begins with scenario setup, where an ego vehicle equipped with AEB logic is placed in a dynamic, orchestrated virtual environment. Other vehicles and pedestrians are added as active participants, and environmental factors like lighting and weather are precisely controlled.

The virtual camera is configured with adjustable FoV, resolution, and shutter parameters. Its outputs — including RGB frames, segmentation masks, and bounding boxes — form the perception input for AEB logic validation. The algorithm is then tested using Time-to-Collision (TTC) metrics, with braking events triggered when collision risk crosses the threshold. Failures such as delayed braking or false detections are logged for analysis.

Edge Cases and Continuous Feedback

The same approach extends naturally into edge case testing – scenarios like a pedestrian emerging between parked cars, vehicle cut-ins during heavy rain, or nighttime crossings under glare. These virtual tests are entirely safe, repeatable, and quantifiable.

Each failure scenario feeds back into the algorithm training cycle. Datasets from missed or delayed detections are used to retrain the perception model, which is then revalidated in the same controlled virtual environment. This closed-loop feedback mechanism continuously enhances AEB performance and robustness.

ISO 26262 and Euro NCAP Alignment

Ansys AVxcelerate and Ansys SCADE together provide a fully traceable, safety-compliant workflow.

This ensures that OEMs achieve regulatory compliance while accelerating validation cycles and minimizing risk.

Conclusion:

As the automotive industry accelerates toward autonomy, virtual validation is no longer optional — it’s essential. By combining Ansys SCADE’s ISO 26262-certified model-based software development with Ansys AVxcelerate’s physics-based simulation, OEMs can bridge the gap between design and reality.

This integrated workflow empowers engineers to validate safety-critical AEB systems with speed, precision, and confidence — ensuring every vehicle on the road can not only think faster but also brake smarter. To learn more in detail you can check out our webinar on AEB system.