Minnano Taxi?-?S.RIDE / Demand prediction
Minnano Taxi (Everybody’s Taxi) Corporation is a joint venture of Sony Corporation, Sony Payment Services, and five Tokyo taxi companies with a combined fleet of nearly 10,000 vehicles. Through innovation including the S.RIDE app for taxi users and a system for taxi drivers that predicts demand, we’re applying AI and other technologies to enable safe and secure transportation as we address issues faced by cities and by places with graying populations.
S.RIDE: Simple, smart, and speedy taxi-hailing
Hashimoto: It’s a common situation: Imagine you’re in a city on a weekday morning when it starts to rain. You try to catch a cab only to find that it is not so easy. Or if you’re older and no longer have a license but live in a rural area where people depend on their cars, you may find it hard to go shopping for what you need every day. With these issues in mind, Minnano Taxi is now taking steps to provide safe, secure, and efficient experience applying Sony AI, information technologies, and business insight.
The first wave of these efforts is S.RIDE, an app for hailing taxis from a smartphone. This service for taxi users was launched in April 2019, initially in Greater Tokyo. Sony designers and engineers have made the UI so simple that all it takes is a single swipe to call nearby taxis. The app also simplifies payment after you reach your destination. You can pay the smart way?-?online, with a credit card registered to the app?-?and you’re out the door. We will be improving services as we act on feedback from users and drivers.
Using AI to predict demand and find where taxis are needed
Migita: While offering S.RIDE to taxi users, Minnano Taxi is also starting to introduce technology that may help taxi drivers and streamline their work. Toward this end, the first step is a demand prediction service that was jointly developed with the Sony R&D team. The service informs drivers of when, where, and about how many passengers are expected within finely subdivided areas around the city. The prediction system was built by applying machine learning to the probe data of participating taxi companies, which have status data for nearly 10,000 vehicles. Meanwhile, another challenge was deciding how to show these results and inform taxi drivers. Feedback from drivers through repeated demonstration testing has greatly improved the system.
The app we developed informs drivers not only about where many passengers are expected to be, but also about factors that affect taxi demand, such as whether it’s raining, what the conditions of public transportation are like, and if any local events are planned. Of course, this information is also used in machine learning to predict demand. We ensured that drivers can also check where they’re more likely to find long-distance passengers, and how many idle taxis are nearby. The response from drivers who have used the system has been positive.
Changes in transportation, including Mobility as a Service (MaaS)
Hashimoto: After October 1 this year, taxi users can take advantage of a new service in Japan that lets them check predetermined fares in an app before boarding. The service may increase ridership, because passengers will feel more comfortable knowing in advance how much they’ll pay, without worrying about higher fares due to traffic or detours. Naturally, we plan to make S.RIDE compatible with the service. Also under consideration is waiving a current ban on taxi ridesharing. There will probably be closer collaboration between taxis and trains, buses, and other forms of public transportation and more cases of consolidated transportation of people and cargo. As AI advances, transportation is poised to evolve greatly over the next five to ten years.
Migita: It was shown through demonstration testing of demand prediction with taxi drivers that this system can increase sales. Years ago, one couldn’t become a taxi driver without a solid understanding of roads and routes. Today, navigation systems have opened the door for many people to become drivers, even if they’re less familiar with some roads. We can guess that, with support from technologies such as AI and sensing, driving can be safer and more secure, and this business can be more efficient. Transportation will surely be changing significantly from advanced technology.
Toward more efficient transport and new services
Migita: Actually, taxis currently spend most time on the road without passengers. We think the demand prediction system will reduce this idle time, making taxis more efficient.
What’s more, the technology can help optimize transportation in general. Applications can extend to car and bicycle sharing, public transportation such as trains and buses, and more. People will be able to get around more smoothly and efficiently in society.
Another potential application is in consolidated transportation, with cargo being carried besides people. In these ways, what we’re addressing are issues faced by cities and by places with graying populations.
Hashimoto: Sony has wonderful imaging and sensing technologies. Equipping taxis with these technologies may support safe driving and next-generation mobility services. From the standpoint of social infrastructure, the information obtained by taxis is truly valuable. Taxis are across the city at all hours, and image analysis of the data from their sensors supports safer driving, and in turn, helps reduce accidents. The data can also be applied in mobility services.
No discussion of the future of transportation would be complete without mentioning autonomous vehicles. For example, imagine autonomous, AI-supported vehicles going around a city, between homes, transit stations, and public facilities as they safely, reliably, and efficiently carry people and cargo to their destination. Maybe we’ll see these mobility services in the near future.