Saving gasoline with the Google Prediction API

Researchers from Ford began testing a new feature of the computer control system, which makes reasonable assumptions about the future route of the car and optimizes the settings. The choice of the mode of the engine in some cases can significantly reduce fuel consumption.

Prediction of the route is based on the Google Prediction API cloud service. The screenshot shows on the left how the probability of each route is calculated.

The system keeps an archive of all vehicle movements and a list of its routes. On the basis of this information, a predictive model is built. The car constantly sends its current coordinates and time to the Google cloud wirelessly. In response, the API returns a list of routes with a probability for each of them.
')
The system is being tested on a Ford Escape SUV with a hybrid powertrain. Since the car has both an internal combustion engine and an electric one, it is important to know which one is preferable to switch on and at what point. Existing hybrids approach the question "stupid" and change the mode of operation depending on the speed of the car. During acceleration - the electric motor, with increasing speed - the engine.

If you know the proposed route, then you can, for example, not turn on the engine during acceleration, if you know that after a hundred meters the intersection will inevitably have to reduce speed. On such "trifles" a considerable amount of fuel is saved.

The Google Prediction API service was launched in August last year, but is only now open for use by everyone (this is a paid service with uptime guarantee).