Vertically Integrated Projects - Autonomous Motorsports Purdue

Get Started

AMP Student Club & VIP

The AMP student club was founded with a vision of building a strong research program for autonomous racing at Purdue. We are currently competing in the Indy Autonomous Challenge, and our goal is to optimize a high speed (average 100 mph) autonomous racing algorithm for a simulated environment. In Fall 2020, AMP joined the VIP program that provides an opprtunity for undergraduate student members to earn academic credit while engaging in AMP's extended research and design projects. AMP is also partnering with the United States Military Academy, West Point, and a professional racing team, whose expertise is providing valuable guidance.


Electrical sub-team is implementing serial parsing microcontroller system to operate the steering, throttle, and braking functionality of the go-kart. The sub-team has successfully integrated the entire electrical subsystem. Integration testing was conducted using a custom javascript-based simulator to send control packets as a stand-in for our navigation computer. The ultimate goal of the electrical sub-team is to have a fully autonomous go-kart that can navigate any track environment at maximum speed.


Software sub-team is implementing the software stack that is centered around Robot Operating System (ROS). The sub-team develped the solution to online path planning and SLAM problems for the first lap of the race. After the development of the software stack, a simulated environment was generated to test the online path planning algorithm on team-designed tracks. Following the successful completion of a lap on various simulated test tracks, preparations are underway to conduct initial physical testing.

Navigation 1

Navigation sub-team 1 is utilizing deep learning techniques to build a camera-based autonomous vehicle capable of racing at high speeds. The sub-team created a realistic environment in the Unity simulator to develop accurate models for real world navigation. Under a supervised learning approach, the car is first driven by hand or the PD controller to generate the training data. Images taken from the car’s perspective are then used to train the model for autonomous mode. Currently, the work is in the model improvisation and data generation phase.

Navigation 2

Navigation sub-team 2 is working on to make the self-driving car faster and more reliable at high speeds. The sub-team is using AirSim, a well-known car simulation plugin for the Unreal Engine, to tackle tasks like PID control of a high speed car and dynamic trajectory generation for overtaking slow moving and cooperative cars nearby while minimizing jerk. The sub-team is developing autonomous vehicle algorithms that pushes the limits of the car which can provide important insights about how to make self-driving cars faster and safer.

Mechanical 1

Mechanical sub-team 1 is working on to simulate vehicle dynamics at high speeds to predict faster race lines and safer trajectories. The sub-team is focusing on lateral vehicle dynamics by using a bicycle model assumption to simplify the dynamics of a vehicle and used MATLAB to compute vehicle response at given speeds and yaw rates. Current work includes looking into more advanced and accurate methods of modeling vehicle dynamics to produce more applicable results, as well as further research on the effects of high speeds on existing models.

Mechnical 2

Mechanical sub-team 2 is developing a program that models the throttle percentage of a high speed vehicle. The sub-team focuses on the longitudinal dynamics of the vehicle and has developed a model that uses RPM and torque of the vehicle to calculate the throttle percentage, using data from an internal combustion engine. Current work includes energy management strategies for hybrid vehicles, tire management during a race, and powertrain control of a high-speed race vehicle.