LoPECS: A Low-Power Edge Computing System for Real-Time Autonomous Driving Services

Jie Tang,Shaoshan Liu,Liangkai Liu,Weisong Shi,Bo Yu

doi:10.1109/access.2020.2970728

Abstract

To simultaneously enable multiple autonomous driving services on affordable embedded systems, we designed and implemented LoPECS, a Low-Power Edge Computing System for real-time autonomous robots and vehicles services. The contributions of this paper are three-fold: first, we developed a Heterogeneity-Aware Runtime Layer to fully utilize vehicle's heterogeneous computing resources to fulfill the real-time requirement of autonomous driving applications; second, we developed a vehicle-edge Coordinator to dynamically offload vehicle tasks to edge cloudlet to further optimize user experience in the way of prolonged battery life; third, we successfully integrated these components into LoPECS system and implemented it on Nvidia Jetson TX1. To the best of our knowledge, this is the first complete edge computing system in a production autonomous vehicle. Our implementation on Nvidia Jetson demonstrated that it could successfully support multiple autonomous driving services with only 11 W of power consumption, and hence proves the effectiveness of the proposed LoPECS system.

Highlights

Many major autonomous driving companies, such as Waymo and Baidu, are engaged in a competition of designing autonomous vehicle which can operate reliably in an affordable cost, even in the most extreme environments
We demonstrated that we could successfully support multiple autonomous driving services with only 11 W of power consumption, and proving the effectiveness of the proposed LoPECS system
AUTONOMOUS DRIVING SERVICES Before going into the details of the LoPECS system design, let us briefly examine the services needed in autonomous vehicle systems

Summary

INTRODUCTION

Many major autonomous driving companies, such as Waymo and Baidu, are engaged in a competition of designing autonomous vehicle which can operate reliably in an affordable cost, even in the most extreme environments. To make autonomous driving universally adopted, the major challenge is to simultaneously enable kinds of computation intensive task on a low-power edge computing system with an affordable price Those autonomous driving services like real-time localization through Simultaneous. The design of such a low-power edge computing system is extremely challenging These computationintensive services are made of complex pipelines and always have tight real-time requirement. At times computing offloading from vehicle to cloudlet leads to energy efficiency, but whether to offload and how to offload remains an unsolved problem To address this problem, we developed a vehicle-edge coordinator to dynamically offload tasks to edge cloudlet to optimize user experience in autonomous driving, in terms of lower power and extended battery life.

AUTONOMOUS DRIVING SERVICES

Findings

CONCLUSION