The Car and The Cloud: Automotive Architectures for 2020

Loading...
Thumbnail Image
Penn collection
Real-Time and Embedded Systems Lab (mLAB)
Degree type
Discipline
Subject
automotive systems
vehicular networks
cyber physical systems
traffic congestion
cloud computing
Computer Engineering
Electrical and Computer Engineering
Mechanical Engineering
Funder
Grant number
License
Copyright date
Distributor
Related resources
Contributor
Abstract

Three trends are emerging in drivers’ expectations for their vehicle: (1) continuous connectivity with both the infrastructure (e.g., smart traffic intersections) and other commuters, (2) enhanced levels of productivity and entertainment for the duration of travel, and (3) reduction in cognitive load through semiautonomous operation and automated congestion-aware route planning. To address these demands, vehicles should become more programmable so that almost every aspect of engine control, cabin comfort, connectivity, navigation, and safety will be remotely upgradable and designed to evolve over the lifetime of the vehicle. Progress toward the vehicle of the future will entail new approaches in the design and sustainability of vehicles so that they are connected to networked traffic systems and are programmable over the course of their lifetime. To that end, our automotive research team at the University of Pennsylvania is devel- oping an in-vehicle programmable system, AutoPlug, an automotive architecture for remote diagnostics, testing, and code updates for dispatch from a datacenter to vehicle electronic controller units. For connected vehicles, we are implementing a networked vehicle platform, GrooveNet, that allows communication between real and simulated vehicles to evaluate the feasibility and application of vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication; the focus in this paper is on its application to safety. Finally, we are working on a tool for large-scale traffic congestion analysis, AutoMatrix, capable of simulating over 16 million vehicles on any US street map and computing real-time fastest paths for a large subset of vehicles. The tools and platforms described here are free and open-source from the author.

Advisor
Date Range for Data Collection (Start Date)
Date Range for Data Collection (End Date)
Digital Object Identifier
Series name and number
Publication date
2012-11-01
Journal title
Volume number
Issue number
Publisher
Publisher DOI
Journal Issue
Comments
@article{carcloud12, author = {R. Mangharam}, title = {The Car and The Cloud: Automotive Architectures for 2020}, journal = {The Bridge on Frontiers of Engineering, National Academy of Engineering}, year = 2012, volume = 42, pages = {25--33}, number = 4}
Recommended citation
Collection