August 27, 2011 at 12:12SignalGuru Traffic Light Prediction Software
Specialists from the Massachusetts Institute of Technology, together with colleagues from Princeton, conducted an interesting study of how an ordinary smartphone can reduce fuel consumption in a car by as much as 20%. Their idea is that the smartphone is equipped with special SignalGuru software , connects to the car computer and is placed in front of the windshield to track the road in front of the car.
The SignalGuru program on a smartphone monitors the turn-on time of traffic signals and gives a signal to the driver in advance - when you need to speed up to slip into the green light, and when you need to slow down in advance, because the green signal will soon turn yellow. Such intelligent predictions can save a huge amount of fuel in the urban cycle, because the driver stops braking and the engine is idling less.
SignalGuru testing was conducted in American cities, where the duration of traffic signals has a fixed time (the so-called "dull" traffic lights), as well as in Singapore, where it depends on traffic intensity. In the first case, the average error was 0.65 seconds, in the second case, a little more than a second. According to scientists, even prediction with an error of 2.5 seconds gives significant fuel savings.
The accuracy of predicting traffic signals depends on the number of cars that install the SignalGuru program. In this case, they can exchange information through a central database.
The SignalGuru program is a much cheaper solution than the so-called “car Internet”, which involves installing radio transmitters in every car and every traffic light on the road.
The program can also work together with a car navigation system. Perhaps in the future, the navigator will be able to lay a route around the city, taking into account traffic signals. Maybe the car itself will even slightly add gas or slow down if necessary, helping the driver on the road.
The SignalGuru program on a smartphone monitors the turn-on time of traffic signals and gives a signal to the driver in advance - when you need to speed up to slip into the green light, and when you need to slow down in advance, because the green signal will soon turn yellow. Such intelligent predictions can save a huge amount of fuel in the urban cycle, because the driver stops braking and the engine is idling less.
SignalGuru testing was conducted in American cities, where the duration of traffic signals has a fixed time (the so-called "dull" traffic lights), as well as in Singapore, where it depends on traffic intensity. In the first case, the average error was 0.65 seconds, in the second case, a little more than a second. According to scientists, even prediction with an error of 2.5 seconds gives significant fuel savings.
The accuracy of predicting traffic signals depends on the number of cars that install the SignalGuru program. In this case, they can exchange information through a central database.
The SignalGuru program is a much cheaper solution than the so-called “car Internet”, which involves installing radio transmitters in every car and every traffic light on the road.
The program can also work together with a car navigation system. Perhaps in the future, the navigator will be able to lay a route around the city, taking into account traffic signals. Maybe the car itself will even slightly add gas or slow down if necessary, helping the driver on the road.