Learning a CNN-based End-to-End Controller for a Formula SAE Racecar

TL;DR

This paper develops CNN-based end-to-end control models for a Formula SAE racecar, demonstrating high accuracy and real-time performance for steering, brake, and throttle prediction, with tools for understanding model behavior.

Contribution

It introduces CNN models for discretized and real-value steering, along with a control network for brake and throttle, and provides benchmarking and visualization tools.

Findings

High accuracy in discretized steering prediction

Low error in real-value steering model

Real-time performance on GPU hardware

Abstract

We present a set of CNN-based end-to-end models for controls of a Formula SAE racecar, along with various benchmarking and visualization tools to understand model performance. We tackled three main problems in the context of cone-delineated racetrack driving: (1) discretized steering, which translates a first-person frame along to the track to a predicted steering direction. (2) real-value steering, which translates a frame view to a real-value steering angle, and (3) a network design for predicting brake and throttle. We demonstrate high accuracy on our discretization task, low theoretical testing errors with our model for real-value steering, and a starting point for future work regarding a controller for our vehicle's brake and throttle. Timing benchmarks suggests that the networks we propose have the latency and throughput required for real-time controllers, when run on GPU-enabled hardware.