This presentation was made at the NAFEMS Americas "Creating the Next Generation Vehicle" held on the 14th of November in Troy.

The automotive engineering community is now confronting the largest technology transformation since its inception. This includes the electrification of powertrains for more efficient consumption and cleaner emissions, the reinvention of the battery with fast wireless charging capabilities and finally the advent of a fully autonomous vehicle. Compounding to these technology changes, the automotive companies design verification process is moving away from a major reliance on physical testing to almost a full virtual simulation product verification process.

The automotive engineering community is now confronting the largest technology transformation since its inception. This includes the electrification of powertrains for more efficient consumption and cleaner emissions, the reinvention of the battery with fast wireless charging capabilities and finally the advent of a fully autonomous vehicle. Compounding to these technology changes, the automotive companies design verification process is moving away from a major reliance on physical testing to almost a full virtual simulation product verification process.

Resource Abstract

Autonomous Transportation is becoming a reality due to rapid development and improvement in the sensing and control technology. Manufacturers are rushing to achieve level 2 autonomous capabilities in not only passenger vehicles, but also in semis, commercial vehicles and construction and mining vehicles.



Autonomous vehicles are complex systems involving interactions between hardware and software in real time. To achieve level 2 autonomy, manufacturers need to develop and test complex driver assistance systems in number of real-world scenarios. This involves these systems being tested for billions of miles in real world scenarios.



To enable efficient development of these systems it becomes important to move to virtual testing. Simulation tools for Autonomous Driving enables achieving efficiency in these tests and ensuring validation and certification for the ADAS algorithms. Further simulation gives you the flexibility to test variety of assistance systems like AEB, ACC, ALC, ESC in number of driving scenarios and conditions in a short period of time.



Simulation helps remove a substantial hurdle in achieving a rigor in the testing to make the ADAS algorithms safe and reliable. Having said that, the sheer number of driving scenarios, driving conditions and systems to be tested make it a mathematically impossible task to achieve. This introduces the need to use mathematical design exploration techniques and machine learning techniques to run simulations on identified worst case scenarios and help sweep through relevant driving situations in an efficient yet effective manner. Further to replicate real world uncertainties in the driving incidents, stochastic analysis of simulations become important to test the robustness of the ADAS algorithms.



In this case study we take an example of a NHTSA defined AEB (Automatic Emergency Braking) Scenario and simulate it within a simulation software. The using an optimization platform we vary parameters such as vehicle speed, distance of the object from the vehicle in an efficient manner to cover a wide range of testing scenarios. We further test the robustness of the AEM algorithms by considering stochasticity on certain key scenario conditions. With this approach we highlight the ability to use mathematical design exploration techniques in conjunction with driving simulations to achieve efficient testing of ADAS algorithms.