SAVS(Smart Access Vehicle Service)

SAVS is a service that that combines the advantages of taxis (on-demand transportation) and fixed route buses (ride-share transportation) and carries out real-time ridesharing vehicle dispatch calculation, by AI. The cloud-based AI platform communicates with smart devices, and, in the constantly changing movement situation of vehicles and people, instantaneously decides vehicle running routes that simultaneously fulfil all spatial movement and desired times. By this technology, we realize optimum transportation at city level.


AI real-time ridesharing

Based on people’s travel demands and the operation status of all vehicles, the AI judges the most efficient vehicle dispatch and automatically sends a pick-up order to the driver. By means of real-time vehicle dispatch determination that allows for passengers to share rides, vehicle dispatch requests from new passengers can be responded to even while transporting other passengers. SAVS enables us to reduce the number of passenger-less vehicles and draw out optimum transportation efficiency with a limited number of vehicles.

SAVS vehicle dispatch image


Effects that can be expected from introducing SAVS

Communal riding

A phrase corresponding to the general communal passenger vehicle operation industry (communal-ride bus industry) governed by the Road Transportation Law. Communal riding indicates that passengers carry out transportation requests and payment individually, and that multiple passengers are made to ride in a single vehicle. In the communal-ride bus industry, there are fixed-route fixed-time buses, fixed-route un-scheduled buses, and regionally operated buses.

Combines riding

In the normal taxi industry, communal riding is not permitted, but it is possible for travelers to form a group before getting into a taxi and then each get out separately. We call the act of using a taxi together in a group ‘combined riding’.

AI real-time ridesharing

SAVS aims to increase user-friendliness for all passengers and make all vehicles more efficient. As a result, sometimes one group of passengers monopolize a vehicle from when they get in to when they get out, and at other times there is pattern of get in→get in→get out→get in… meaning that the situation in which somebody is riding continues for a long time. This cannot be explained only in terms of the existing concepts of ‘communal riding’ or ‘combined riding’. Therefore, we decided to called SAVS an AI real-time ridesharing transportation service. When passengers are given rides together, if their directions of travel are too different it is not ‘incidental’, nor is it ‘incidental’ if there is too much time between rides. Furthermore, the feeling of travel being ‘incidental’ varies for each passenger. SAVS uses AI to realize sophisticated real-time ridesharing.


Application screen image

SAVS passenger application

This is a passenger-oriented Web application, optimized for use on smartphones. When you send information such as pick-up location, drop-off location, number of people, desired time etc., you will instantly be notified of the vehicle dispatch result, including vehicle information, estimated pick-up time, etc. After confirming vehicle dispatch, you can check the current location of your assigned vehicle on the map.

SAVS driver application

This is the driver-oriented application. The vehicle’s current location, the next destination and future scheduled destinations, designated by AI, are displayed on the map. By having each driver perform repeated pick-ups and drop-offs, in accordance with the instructions on the screen, we realize optimum transportation throughout an entire city. Drivers use the dedicated application by installing it on a commercial Android tablet.

SAVS call-center application

The call center comprehends and manages the vehicle dispatch situation, vehicle operation situation, and the overall situation of the system. Through management by call center (equivalent to a taxi dispatch center) staff, vehicle dispatch bookings can be made on behalf of customers. This application can be used with a computer and Web browser (Chrome is recommended).


Basic system structure

① SAVS applications

Provide a basic user interface for using SAVS to provide services, in the form of the passenger app., driver app., call center app. and so on. There is no need to develop or customize the applications, so service can be initiated quickly and at low cost.

② Vehicle dispatch server

Carries out every kind of data handling needed to realize the service that uses the AI dispatch algorithm, such as management of dispatch booking data and dispatch result data, management of users, fixation of running vehicle schedules etc. Input data via the API is transferred to the engine layer, then the engine layer returns the processing result via the API. We make maximum use of the features of the cloud service to provide a platform equipped with sufficient flexibility and resistance to infrastructural impairment to conform to the volume of access.

③ Dispatch calculation engine

Based on road network data of the relevant area, AI dispatch calculation is carried out by the combination of a unique algorithm and every kind of dispatch parameter. The engine deals with dispatch calculation for a wide variety of on-demand transportation, by means of hyper-fast real-time dispatch calculation (result output in under one second) and batch dispatch calculation focused on efficiency.

④ Simulator

Carries out, by simulation, measurement of the effect of introducing SAVS, based on virtual travel demands, person-trip survey data, and actual travel transportation data etc. Simulation is useful for formulating plans in response to the area or purpose of SAVS introduction, because, from the balance of number of travel demands and number of provided vehicles, we can obtain an approximation of the average waiting time or trip time and driving distance etc.

⑤ API link service

All the SAVS functions on the cloud are provided by a REST API. By using the necessary API in accordance with procedure, SAVS can be incorporated into a unique application or IoT device. This helps to improve the quality and efficiency of services combined with AI vehicle dispatch, such as a MaaS app., tourism app., delivery system, and so on.