Monday, August 29, 2011

#WIRELESS: "Social Net Saves Fuel With Smartphone"

Crowd-sourcing the state of traffic-lights from dash-mounted smartphones enables smart social networkers to keep cars rolling through green lights, thus cutting fuel consumption by up to 20 percent.

Smarter social networking apps could actually save you money on fuel, according to the Massachusetts Institute of Technology and Princeton University, which recently demonstrated its SignalGuru app. The app watches for red lights and then alerts friends in cars behind the driver to adjust their speed to avoid it.

The idea behind the MIT/Princeton research project is to crowd-source red-light timing information and then share that information with other drivers. The smarter software works by dash-mounting a smartphone so that its camera is facing slightly upwards where it can see the traffic lights as a car drives past them. A cloud-based server then collects each traffic light's state and enters it into a real-time database. Then, depending on the current location of a particular car, drivers will be informed as to the proper speed to drive in order to hit the next light while it is green. As a result, SignalGuru users were able to save an average of 20 percent on their fuel costs compared to those who had to stop at lights.

Where previous experimental traffic-light advisory systems used GPS data or data from traffic sensors, SignalGuru uses visual data from cellphone cameras. (Source: MIT)

The main investigator on the project, MIT doctoral candidate Emmanouil Koukoumidis, claims that 28 percent of energy consumption and 32 percent of carbon dioxide emissions in the United States are from automobiles, which should enable SignalGuru to have a strong, positive impact on both the environment and the economy if enough people start using commercial versions of the app.

Koukoumidis was inspired to create the app by the growing trend of drivers that are already dash-mounting their smartphones in order to use them as navigational devices. SignalGuru just repurposes the forward-facing camera to keep an eye out for traffic lights. Eventually SignalGuru could also be integrated with these navigation apps, in order to offer not only a recommended speed, but perhaps even route changes that make sure the driver hits green lights all the way.

Koukoumidis, who worked with MIT professor Li-Shiuan Peh and Princeton professor Margaret Martonosi on the project, also envisions crowd-sourcing other related information as drivers proceed on their commutes. For instance, advising as to where the lowest gas prices are available along a driver's route, where parking spaces are available, and even tracking the progress of city busses for smartphone users who are hoofing it.

SignalGuru worked best in Cambridge, Mass., where traffic lights operate on a fixed schedule, allowing drivers to synchronize their driving speed with the changing traffic lights to within two-thirds of a second. The system worked less well in Singapore tests where traffic lights track how many cars are on the road and vary their timing. Nevertheless, SignalGuru was able to keep Singapore commuters driving through green lights with and error of slightly more than a second, which still enabled drivers to reap significant fuel savings.

Further Reading