RoboRacing, our team focused on autonomous racing robots, has been making platforms of various shapes and sizes over the 6 years since its inception. Initially formed to compete in the International Autonomous Robot Racing Competition (IARRC), the first platforms were approximately 1/10th scale Remote Control (RC) cars with laptops mounted directly on the chassis. While the team still competes in IARRC, the team also develops platforms on a larger (sometimes human-sized) scale. The first larger scale car was Bigoli, an approximately 1/3rd scale platform that was designed to compete at the Sparkfun Autonomous Vehicle Challenge (AVC) starting in 2017. Unfortunately, Sparkfun was canceled following the 2018 season. This led to the end of the development of Bigoli, but it was certainly not the end of larger scale autonomous platforms for RoboRacing.
Following the cancellation of Sparkfun AVC, the team searched for a new competition and discovered the Electric Vehicle Grand Prix (evGrand Prix). Originally started as a manually driven electric go-kart competition (which still serves as a major part of the event), a new autonomous division was started around 2018 (what a fortunate coincidence!). The team was happy to find a great successor to Sparkfun AVC!
The premise of Autonomous evGrand Prix is to take a Top-Kart frame (a frame manufacturer typically focused on gas-powered go-karts) and make it electric powered and autonomous. The competition, held in the infield of the Indianapolis Motor Speedway, requires navigating other vehicles and barriers as fast as possible in a head-to-head race around the course.
In order to race around the competition, we need a kart to make autonomous. Introducing RoboRacing’s latest pasta-named robot, Rigatoni!
The competition requires teams to use some standard components. This puts the focus of the competition on autonomous decision making and navigation, supported by a modified mechanical and electrical platform. The base frame of Rigatoni is a Top Kart USA EAV Chassis. The chassis includes the original aluminum tubing, base plates, drive axle, steering system, and braking system. There are four 12V marine batteries connected for a 48V base system, which powers the provided motor controller, DC motor, and safety contactors. Everything else is up to the team to decide!
The team originally began work on Rigatoni back in the Fall of 2019. Though the team had the aforementioned standard components, none of the CAD was provided. The mechanical team has been working on modeling the base CAD of the frame and auxiliary components in Autodesk Inventor. This process has led the team to creating new custom components to interface with the existing system to enable autonomous control. For example, a steering assembly was designed that allows electric control using the on-board computer while also allowing manual steering for moving the car when it is not powered. After the assembly is designed, it needs to be manufactured. Preliminary mounting systems for the camera, GPS, and power electronics have already been installed and will be improved over the year as needed.
The RoboRacing electrical team has been making steady progress as well. There are 5 custom printed circuit boards (PCBs) that handle steering, driving/braking, Emergency Stop (1 for wireless, 1 for interfacing), and remote-control. These various boards need to be both powered and communicate back to the central control computers. wo of these problems have been combined into one solution with the use of one specific implementation of Power over Ethernet (PoE). This version of the PoE standard (Mode B 10/100) has Ethernet communication connectors with spare wires that are typically unused at lower communication speeds. These extra wires can carry 24V power to the custom PCBs while maintaining the reliable and fast communication of Ethernet to interface back with a main control computer. Standard Ethernet communication is becoming increasingly common in modern vehicles, so the application of this technology on Rigatoni directly mirrors industry trends.
Software is what makes the difference between a standard electric vehicle and an autonomous electric vehicle. The software platform, which runs on the Robot Operating System (ROS), has three main components: localization, mapping, and planning and control. Localization involves determining the robot’s position on the track using on-board sensors. Rigatoni utilizes a LiDAR (shoots out lasers to determine the distance of objects), cameras, GPS, an inertial measurement unit (measures acceleration and rotation) and many other sensors to help sense the obstacles and the environment. This information is fed into mapping, which helps determine the location of obstacles and other robots. The final component is planning and control, which involves determining a path to get around obstacles and attempting to travel along the determined path.
The original goal was to compete in evGrand Prix 2020, scheduled to occur in April of 2020. Due to the COVID-19 pandemic, the 2020 competition was canceled and development on the robot had to come to a halt due to campus closure. As of right now, evGrand Prix 2021 is currently planned to occur in late Spring. The team has made a lot of progress so far, and they hope to continue to ride this momentum into the 2020-2021 season!