Asian car giant Toyota is looking to outdo Uber by predicting demand for rides. By utilizing data gathered from smartphones, cab locations, climate conditions, and numerous other factors, Toyota believes they can optimize Tokyo’s entire taxi fleet. The project comes as a collaboration between Toyota, taxi-calling startup Japan Taxi, telecom corp KDDI, and Accenture Plc., and together recently announced a 94% accuracy rate. Though Toyota has invested in Uber, they also directly fund Uber competitors like Japan Taxi.

Analysts say that taxi markets are prone to not meeting demand, whereas ride-share markets ignore common supply depressors like poor weather. Software and variable pricing allow Uber and other rideshare companies to move drivers to where demand is focused, giving at least a 20% sales boost.

Toyota first began equipping 500 Tokyo taxis with their TransLog devices last year. Toyota expects the installation to be fully complete by the end of 2018, streaming real-time data about taxis that should help the program better direct drivers to high-demand locations. Observers also say the AI-powered system should make automated, driverless cars an easier transition.