Imagine merging onto the highway and instead of acting as a driver, you are now the passenger with the car operating safely without the need for human input or oversight. In the automotive engineering industry, this situation is known as Level 4 vehicle autonomy.

Before making this type of vehicle autonomy available to the public, automotive engineers must ensure that the self-driving car meets rigorous industry standards for safety and reliability. As the industry continues to explore ways to safely ensure this type of vehicle autonomy, so do student engineering competitions such as SAE International’s AutoDrive Challenge, in which students experience firsthand the process of designing highly autonomous vehicles with industry-standard software tools like MATLAB and Simulink.

Over a three-year period, the AutoDrive Challenge recreates a real-world engineering environment in which student teams face the same design problems that automotive engineering professionals face as they assess and address the complexities that come with full automation.

The recently completed Year One phase of AutoDrive called on students to master three main challenges. In the first challenge, teams needed to demonstrate that the vehicle can stay in its lane on a straight road and detect traffic signs. The second challenge required the vehicle to stay in its lane on a road with lots of curves, similar to what a driver would experience on a rural road. The third challenge focused on whether the vehicle can detect and avoid objects, similar to highway driving.

In Year Two, which kicks off with a one-day workshop in Ann Arbor, Michigan, on September 14, the focus will shift to urban environments. Students will encounter simulated environments for navigating highways and avoiding pedestrians—issues that automotive engineers are currently tackling. Simulating these more complex environments requires students to develop and fine-tune algorithms and sensors, giving their teams a competitive advantage and producing results that can be applied to real-world automotive designs. The latter helps industry reduce the time and investment it must spend on its own autonomous vehicle research and development.  

With each successive year of competition, teams tackle more advanced aspects of automated driving and begin employing deep learning, artificial intelligence, and other technologies to adapt to increasingly difficult driving challenges. Year Three is scheduled to begin in June 2019.

Simulation Techniques Drive Competition Environments

Building a highly autonomous vehicle requires flexibility and constant adjustment. That’s why simulation tools like MATLAB and Simulink are essential to AutoDrive. At any moment, sensors can be collecting hundreds of thousands of data points that a self-driving car needs to interpret and respond to in order to be fully functional and, more importantly, safe. Simulation helps competitors develop more accurate algorithms by introducing varying environmental conditions as well as other traffic participants and  quickly optimize algorithms.

For example, simulation allows student teams to use 3D environments that more accurately apply sensors, gathers virtual test data, fine-tunes and develops new algorithms, and even offers the opportunity for teams to test drive their cars—similar to a racing videogame. In accessing these simulated environments, teams write and tune algorithms capable of adapting to winter road conditions and crowded city streets using the same technology as industry automotive engineers.

It’s not Just About the Technology

Although learning and understanding the engineering technology needed to develop an autonomous vehicle is a key component of the AutoDrive Challenge, the competition also provides students with the opportunity to build project management and communication skills.

Autonomous cars are new to society, and for adoption of this new technology to occur, the public needs to understand self-driving cars and consumers need to have faith in the end-product they purchase. To that end, AutoDrive student teams actively reach out to their local communities to collect public sentiment regarding autonomous vehicles—everything from the ideal look of the car to safety concerns.

The student teams regularly post on social media channels to help spread social awareness on the engineering technology and foster more adoption of vehicle autonomy.