Laboratory Notes: CarTel – a software system for data collection from moving cars

Samuel Madden, Hari Balakrishnan
Computer Science and Artificial Intelligence Laboratory

Millions of Americans spend an hour or more a day in their cars. For such users, there is an obvious benefit to technology that can improve the quality of their drive, for example by reducing congestion. Similarly, such cars provide an opportunity to exploit their inherent mobility to collect data about traffic conditions, wireless coverage, and environmental data such as road and air. In the CarTel project, we are developing software to address this challenge and exploit these opportunities.

This cartel screen shot shows a prototype interface with a user visualization of a recently driven route.

This cartel screen shot shows a prototype interface with a user visualization of a recently driven route.

A CarTel equipped car is set up as a mobile sensing platform for measuring traffic, road surface quality, pollution, and car status and for diagnostics information. The primary research involves building wireless network protocols and data management software for a world where nodes are mobile (e.g., on cars), and media-rich sensor data is collected using a variety of sensors (e.g., cameras, vibration sensors, On-Board Diagnostic (OBD-II) data from the car’s internal sensor network, Wi-Fi and cellular network monitors, etc.). Furthermore, rather than assuming cars will have an always-on Internet connection, in CarTel we are developing opportunistic networking software that allows cars to send and receive data even though they are only intermittently connected to servers on the Internet (e.g., when they drive by a WiFi access point.)

Two example applications of this platform include:

> Traffic monitoring and route planning. Every CarTel node reports its (anonymized) location once per second using GPS. We are using this information to develop statistical models of traffic delays at various times of day on different road segments to perform delay-aware route planning.
> Civil and environmental monitoring. When equipped with additional sensors to detect vibration and other conditions, cars can act as excellent “probes’’ to record environmental and road conditions. For example, we have performed an initial study to identify the location, frequency, and magnitude of potholes in the Boston area using accelerometers in a fleet of taxicabs equipped with CarTel hardware and software.

To realize these applications, CarTel cars are outfitted with a small, Linux-based PC and an array of mbedded sensors. Data is collected by an embedded database system and relayed via our opportunistic networking software to a centralized portal website where data can be browsed, analyzed, and visualized. The screen shot (above) shows a prototype interface with a user visualization of a recently driven route.

For more information see the CarTel website.

One Response to “Laboratory Notes: CarTel – a software system for data collection from moving cars”

  1. Amy says:

    This type of technology will be most important in the future. We have more and more cars on the road, creating terrible traffic jams. With a properly worked out route that sends different individuals in different paths, congestion can be dealt with to some extent.

Leave a Reply

Replies which add to the content of this article in the EECS Newsletter are welcome. The Department reserves the right to moderate all comments. If you would like to provide any updated information please send an email to newsletter@eecs.mit.edu.