On May 28th and 29th the Grand Cooperative Driving Challenge (GCDC) 2016 took place at the Automotive Campus in Helmond, The Netherlands. Of the ten competing student teams the team from Halmstad University won the race with self-driving vehicles which communicated with each other and the infrastructure.
As a technology sponsor of the GCDC 2016 event and i-GAME project TASS International facilitated several student teams with the simulation platform PreScan for testing their automated vehicles in the various traffic scenarios. During the GCDC 2016 event the student teams were assessed on three different Automated Driving traffic scenarios: automatic merging on the highway, cooperatively crossing an intersection and automatically making room for an emergency vehicle on a highway.
The winning team of Halmstad University – consisting of members Viktor Frimodig, Oscar Uddman Jansson, Jérôme Detournay, Thomas Rosenstatter, Golam Shahanoor, Wojciech Mostowski, Víctor Díez Rodríguez and Maytheewat Aramrattana – are all students studying the Master’s Programme in Embedded and Intelligent Systems at Halmstad University (except for two PhD students) and have surely had a meaningful and fun experience in the area of Cooperative Intelligent Transport Systems and Cooperative Automation.
Giving the student team some time to cool down after many weeks of preparations and long working days TASS International followed up with the team to discuss their feedback about the PreScan software the team used during their GCDC preparations.
Víctor Díez Rodríguez from team Halmstad University explains: “We gave PreScan two main uses: to test the car controller and to test the interaction protocols for GCDC. Testing the controller was more or less straightforward. Both PreScan and the implementation of our controller use Simulink, making it easy to interface the vehicle information from PreScan to our model. The higher level control of our system was written in Java, so testing the interaction protocols was a little bit more difficult. We managed to run our complete implementation for several vehicles, dedicating to PreScan the task of simulating the vehicle dynamics and the output of the sensors of the car.”
“PreScan helped us to find bugs in our implementation in early stages of the project without spending too much time in the car, which speeded up quite a lot the development cycles. We took advantage of PreScan's capabilities to load real maps and use GPS coordinates to run hardware-in-the-loop tests, for example making it possible to platoon with a simulated car. The first time we had the chance to run the scenarios with other teams was just one week before the competition so, without PreScan, it would have been impossible to make any test concerning the interaction with other vehicles before arriving to Helmond.” says Rodríguez.