Japan's KDDI has collaborated with Toyota Motor Corporation and JapanTaxi to develop a taxi dispatch system that predicts demand for taxi services, and have commenced a trial deployment in metropolitan Tokyo.
The new system uses AI to predict the number of occupied taxis in Tokyo every 30 minutes, by combining data from taxi service logs with predictions of demographics made by location-based big data from smartphones and event information.
The system also uses demographic predictions as well as other factors that affect taxi demand including weather, public transport availability and events at large facilities. During a test run in Tokyo the system demonstrated an accuracy rate of around 94.1%.
This information is then sent to a map on tablet PCs equipped in selected taxis, allowing drivers to position their taxis based on the best balance of supply and demand.
During the test run in February, drivers who used the system recorded an average sales increase of 20.4% on a month-to-month basis, compared to an overall average increase of 9.4%.
The system will next be implemented in dozens more taxis and is expected to be rolled out for widespread use later in the year.
KDDI developed demographic prediction technology for the trial using its location-based big data from smartphones. Toyota is processing and analyzing the data using its Mobility Services Platform, while JapanTaxi is responsible for developing the app and collecting taxi logs and related information.
The collaboration serves as a good example for operators seeking to monetize their telco big data.
KDDI involved in AI taxi dispatch pilot, working with Toyota and JapanTaxi on a system that predicts demand for taxi services