Methodical driver assistance system testing

Objective Evaluation of Driver Assistance Systems: Method and Toolchain in Focus

Driver assistance systems (ADAS) have become an essential part of modern vehicles—but how can their performance be evaluated objectively and transparently? This question was at the heart of a collaboration between MdynamiX, the Institute for Driver Assistance and Connected Mobility (IFM) at Kempten University of Applied Sciences, and auto motor und sport. At the end of last year, auto motor und sport tested four vehicles with a focus on ADAS using a scientifically validated methodology and a powerful toolchain. In this article, we take a closer look at the evaluation process.

The Method: Transparency Through Objective and Subjective Evaluation

To comprehensively assess the quality of driver assistance systems, IFM’s methodology was applied in collaboration with MdynamiX. A key aspect of the approach was combining subjective driving impressions with objective measurement data.

The MXevalApp played a crucial role in this process. It enabled test drivers and experts to digitally, simultaneously, and systematically capture subjective perceptions. This ensured precise documentation of driving experiences while also providing transparency throughout the entire development process—from customer feedback to final engineering analysis.

MXeval and MXevalApp: Capturing, Analyzing, and Optimizing Data

During testing, all vehicles were equipped with measurement technology to ensure detailed, objective data collection. The recorded measurement data was then analyzed using the powerful software tool MXeval.

MXeval enables comprehensive data processing, enrichment with ground-truth information, and the calculation of function-specific KPIs. These key performance indicators (KPIs) are essential for defining precise target values for future vehicle models and advancing the development of driver assistance systems based on data-driven insights.

With real-time data analysis directly in the vehicle, engineers were able to immediately assess system performance and identify areas for improvement—a clear advantage for efficient development.

Conclusion: Efficient Evaluation for Better Driver Assistance Systems

The combination of subjective perception and objective measurement data provides a reliable, transparent, and customer-focused evaluation of driver assistance systems. The innovative toolchain, consisting of MXevalApp and MXeval, enables precise data capture, analysis, and optimization—offering real value for engineers and developers.

The first part of this three-part series about the ADAS-tests has already been published in auto motor und sport (Issue 6/25).

Watch the video now to get all the details!