UM::Autonomy (UM::Autonomy) is an undergraduate student team at the University of Michigan focused on the development of autonomous vehicles for the International RoboBoat Competition. Participation at all stages of our students’ college careers offers a wide array of learning opportunities that come with a multidisciplinary design project.
Since the RoboBoat’s conception in 2008, the associated challenges and tasks have been directly correlated to the maturity of global autonomous vehicles research. As challenges around autonomous vehicles continue to increase in complexity, so does the competition. As a result, the competition has transitioned away from computer vision based challenges and is increasingly dependent on our ability to localize and map our surroundings.
Historically, we localized and mapped surroundings by combining GPS, gyroscope, compass, camera and LiDAR data. However, we were frequently plagued by sensor errors and the limited capabilities of our system. This system was unsustainable. However, with the help of Sparton and through the implementation of the Sparton AHRS-8, we have started to change that.
Sparton AHRS-8 Application
The Sparton AHRS-8 has become increasingly critical to the success of UM::Autonomy since we received it. Due to its high heading accuracy, we have been able to replace our Fiber Optic Gyroscope (FOG). This allowed us to decrease weight, condense our electrical systems, and simplify many parts of our code base. This year, we have also implemented the Sparton IMU into our localization system. In our prior system, we depended purely on a GPS to provide our position, which led to jittery position measurements that caused our boat to navigate incorrectly. By combining our IMU and GPS with a particle filter, we have been able to achieve accurate, smooth localization which enabled us to qualify for finals in the 2018 AUVSI RoboBoat competition.
In the future, we expect to transition our localization system into an Extended Kalman Filter based system, which will be even more dependent on the Sparton IMU than our current system. As we start to become more dependent on core sensors such as the IMU, we are hoping to acquire a duplicate of our critical components. Currently, our development is frequently stagnated by the need for multiple teams to use the boat and our sensors at any one time. A duplicate of our core sensors would give the ability to have one group testing on the water and a second to be troubleshooting alternative AI or electrical problems simultaneously. We also run the risk at competition of having a single electrical failure, eliminating our ability to compete. Enabling the UM::Autonomy team to do year round, multidisciplinary, simultaneous testing would be a game changer. The Sparton AHRS-8 has become essential to our success and we hope to put our members in a position to utilize the full potential of the AHRS-8 and the educational benefit they receive from working with professional equipment.
UM:Autonomy – Rourke Pattullo, Benny Johnson, Ryan Johnston