SoftBank Begins Conducting Autonomous Driving Field Tests to Realize Fully Automated Driving Operations.

Scope includes driving route design and AI remote monitoring in anticipation of Japan’s legalization of Level 4 autonomous driving.

March 10, 2023 – Softbank Corp announced it began conducting field tests in the Takeshiba area of Minato-ku, Tokyo, in January 2023 that verify driving route design, AI remote monitoring, and other tasks to achieve fully automated autonomous driving operations, with the aim of realizing the early implementation of a highly sustainable autonomous driving service in Japan.

When Japan’s amended Road Traffic Act is enacted in April 2023, Level 4 autonomous driving (“High Driving Automation”) vehicles will be allowed to operate on public roads. While there are high expectations for the practical application of autonomous driving, such as solving the driver shortage issue and reducing traffic accidents, providing such services requires many functions and systems, and the cost of implementation and maintenance is considerable. To address the legalization of Level 4 vehicles and the previously mentioned challenges, SoftBank will continue to conduct field tests to achieve the automation of driving operations with the goal of realizing the early implementation of a highly sustainable autonomous driving service in Japan.

Overview of the Field Tests

Softbank Autonomous Driving

Verification of the Autonomous Driving System

With the enactment of the revised Road Traffic Act in Japan, operators will be required to appoint autonomous driving operation supervisors to handle operations and emergency responses when driving under specific conditions. Autonomous driving operation supervisors can respond either in the vehicle or remotely. To realize the automation of operations in the future, SoftBank is conducting research and development on remote operations.

Automation with AI Remote Monitoring

Monitoring autonomous vehicles remotely account for a significant amount of workload involved in autonomous driving operations. It is difficult for a single monitor to ascertain the surroundings of multiple vehicles and respond to changes in real time. To address this challenge, SoftBank is developing an AI-powered remote monitoring system that can automatically acquire and edit necessary information for the monitor to respond, and SoftBank has been conducting verification tests utilizing this AI system.

Development of In-vehicle Operation Support System

In implementing autonomous driving without a driver, the system must handle various tasks that a driver previously handled. SoftBank is developing an operation support system that incorporates multiple functions such as automatically analyzing the situation inside a vehicle using cutting-edge sensing technology to detect the human presence or when a person falls, and then provide voice guidance and coordinate with remote monitors based on the analysis results. SoftBank will continue to develop these systems while integrating them with the AI-powered remote monitoring system for various MaaS (Mobility as a Service) use cases, such as taxis and small microbuses, to ensure appropriate responses even in an environment with a reduced workforce.

Verification of Operation Optimization through Digital Twin

Since autonomous driving uses the Autonomous Driving System (ADS), which learns the driving course and operates the vehicle automatically, it is necessary to establish the Operational Design Domain (ODD). Furthermore, accurate feedback is also necessary to improve the quality of the autonomous driving operation system.

Automation of Route Planning through Simulation

To establish ODD, it is necessary to understand specific environments and conditions, such as traffic conditions and hazardous areas along driving routes. Through the utilization of a Digital Twin virtual space, accident data, traffic data, and pedestrian flow data, SoftBank will continue to efficiently analyze driving routes through simulations and verify the technology for the automatic calculation of optimal driving routes.

Feedback to the Autonomous Driving Operation System

In autonomous driving services, it is important to accurately capture the characteristics of the traffic environment on the road and understand events that occur. The events that occur during the actual operation of autonomous vehicles will be aggregated onto a platform and scenario-based verification will be conducted to support the quality improvement of various autonomous driving operation systems.

Driving Route in the Field Tests

The autonomous driving operation system utilized in this field test is the ADS (Autonomous Driving System) provided by May Mobility, Inc. (“May Mobility”). Based on the data obtained from the ADS, SoftBank will work with May Mobility to develop and verify the functions necessary for autonomous driving operation management and technical requirements for autonomous driving. Safety management for this autonomous driving field test will be conducted in collaboration with MONET Technologies Inc.

SoftBank will hold an event called “ギジュツノチカラ Advanced Tech Show 2023” on March 22 and 23, 2023 to showcase various initiatives of SoftBank’s Research Institute of Advanced Technology, including the verifications mentioned in this press release. For more details about the event, please visit the event website. (* The event will be held in Japanese only)