# Go-Kart

## Overview

This document describes the hardware and software configurations of the autonomous Go-Kart, which uses Autoware for autonomous driving. It serves as a guideline to select hardware and software components to build autonomous vehicles. This document is not a step-by-step instruction or a Q&A book. Where such material exists, this document links to it.

## Design overview

The charts give a high-level overview of the design. The [metrics page](<../../metrics/index.md>) defines most of the axes. It has no definition for ODD, supportability, traffic density, camera, GPS, LiDAR, and radar.

The Go-Kart design scores from 0 to 5 on each of the nine meta-info metrics.

The Go-Kart design scores from 0 to 5 on each of the eleven tech-info metrics.

## Who should read

Read this page if you want to build a small autonomous vehicle and look for a place to start.

## References and resources

- Complete Go-Kart with Autoware documents: [Go-Kart documentation](<https://go-kart-upenn.readthedocs.io/en/latest/>)
- ROS 2 Foxy setup and autonomous code: [`ros2_foxy_purepursuit` branch](<https://github.com/mlab-upenn/gokart-sensor/tree/ros2_foxy_purepursuit>)
- ROS 2 Humble setup and autonomous code: [`ros2_humble_purepursuit` branch](<https://github.com/mlab-upenn/gokart-sensor/tree/ros2_humble_purepursuit>)
- STM32 Nucleo controller code: [`STM32 Control` folder](<https://github.com/mlab-upenn/gokart-mechatronics/tree/main/STM32 Control>)

> [!NOTE]
>
> The Go-Kart code uses ROS 2 Foxy or ROS 2 Humble. Autoware 2.0.0 supports ROS 2 Humble on Ubuntu 22.04 and ROS 2 Jazzy on Ubuntu 24.04. To build Autoware for the Go-Kart, see [Source installation](<../../../installation/autoware/source-installation/index.md>).

## Contributors

- Autoware Center of Excellence (under Prof. Rahul Mangharam)

## Targeted use cases

| Use Cases: | Education | Racing | PoC | Open AD Kit | Commercial Services/Production |
| :---: | :---: | :---: | :---: | :---: | :---: |
| Go-Kart | ☑ | ☑ | ☑ | ☐ | ☐ |

## ODD

| ODD | Cargo | Bus | RoboTaxi | Low Speed (Utility) Vehicle | Delivery Robot |
| :---: | :---: | :---: | :---: | :---: | :---: |
| Go-Kart | ☐ | ☐ | ☐ | ☐ | ☐ |

## Automation level

| Automation Level | 0 | 1 | 2 | 3 | 4 | 5 |
| :---: | :---: | :---: | :---: | :---: | :---: | :---: |
| Go-Kart | ☑ | ☑ | ☑ | ☑ | ☐ | ☐ |

## System architecture

The autonomous Go-Kart is a one-third scale autonomous vehicle for development and prototyping. You can use it to develop and test software components of Autoware. The photo shows the drive-by-wire parts: the drive motor with throttle-by-wire, the upper and lower steer-by-wire, the brake-by-wire, and the back shelf with the main controller.

## Hardware configuration

### Chassis

- TopKart: [bill of materials and purchase links](<https://go-kart-upenn.readthedocs.io/en/latest/Bom.html>)

### Sensors

- LiDAR: Ouster OS1 and Velodyne Velarray M1600
- Camera: OAK-D camera
- GNSS: RTK-GNSS (Septentrio receiver)
- IMU: Adafruit 9-DoF IMU

### Computing platforms

- x86 laptop with NVIDIA GPU
- STM32 Nucleo microcontroller boards (Nucleo L432KC and Nucleo F439ZI)

## Software configuration

The diagram shows how the sensors and the microcontrollers connect to the computer.

The Go-Kart sensors and microcontrollers connect to the computer over Ethernet, USB, and USB-TTL.