Pylot: A Modular Platform for Exploring Latency-Accuracy Tradeoffs in Autonomous Vehicles

TL;DR

Pylot is a modular, high-performance platform designed for autonomous vehicle research, enabling analysis of latency-accuracy tradeoffs and supporting simulation and real-world deployment.

Contribution

It introduces a flexible dataflow-based architecture for AV components, with reference implementations and case studies demonstrating its effectiveness.

Findings

High score on CARLA Autonomous Driving Challenge

Effective integration with simulators and real vehicles

Insights into context-dependent component design

Abstract

We present Pylot, a platform for autonomous vehicle (AV) research and development, built with the goal to allow researchers to study the effects of the latency and accuracy of their models and algorithms on the end-to-end driving behavior of an AV. This is achieved through a modular structure enabled by our high-performance dataflow system that represents AV software pipeline components (object detectors, motion planners, etc.) as a dataflow graph of operators which communicate on data streams using timestamped messages. Pylot readily interfaces with popular AV simulators like CARLA, and is easily deployable to real-world vehicles with minimal code changes. To reduce the burden of developing an entire pipeline for evaluating a single component, Pylot provides several state-of-the-art reference implementations for the various components of an AV pipeline. Using these reference implementations, a Pylot-based AV pipeline is able to drive a real vehicle, and attains a high score on the CARLA Autonomous Driving Challenge. We also present several case studies enabled by Pylot, including evidence of a need for context-dependent components, and per-component time allocation. Pylot is open source, with the code available at https://github.com/erdos-project/pylot.