Connected vehicle technology is emerging as a cornerstone for next-generation mobility systems, particularly in the transition toward autonomous driving. By enabling continuous communication between vehicles, infrastructure, and other road users, such systems promise to reduce accident risks and enhance roadway safety. In Japan, where vulnerable road users such as motorcyclists and pedestrians account for approximately 70% of traffic accidents, the potential impact is significant.
SoftBank Corp.’s Research Institute of Advanced Technology, in collaboration with Honda Motor Co., Ltd. and Honda R&D Co., Ltd., has undertaken a series of technology verifications aimed at predicting collision risks and alerting drivers in real time. Yakumo Fujimoto, of SoftBank’s Next Generation Connected Development Section, has been central to this work since joining the company in 2021. His research spans V2X communication technologies, digital twin applications, and API-based data acquisition and control, now focused on connected vehicle communication systems.
The trials leveraged Cellular V2X, a variant of vehicle-to-everything communication that uses LTE and 5G networks to exchange data between vehicles, infrastructure, and other entities. SoftBank provided the cellular V2X environment and developed a data integration platform capable of aggregating traffic conditions across an entire road network. This platform combined positional and speed data from connected vehicles with information on non-connected vehicles, gathered via roadside sensors installed along the Shin-Tomei Expressway.
One distinguishing feature of the verification was the inclusion of roadside sensor data for vehicles lacking communication capabilities. This allowed the system to track and integrate movements of both connected and non-connected vehicles, a critical capability during the transitional phase when mixed fleets share the road. Honda contributed a “risk algorithm” that predicts driver behavior several seconds ahead, using real-time data streams from the integration platform.
The demonstration was part of the “Road-Vehicle Cooperation Demonstration Experiment for the Autonomous Driving Era on Expressways,” supported by Central Nippon Expressway Company Limited (NEXCO Central). A representative use case simulated a scenario in which a motorcycle, due to driver inattention, abruptly changes lanes on the expressway. In this situation, the motorcycle’s proximity to a non-connected vehicle ahead and its failure to detect a connected vehicle in the adjacent lane could lead to either a collision or sudden braking by the neighboring vehicle.
By applying Cellular V2X over both 4G LTE and 5G networks, SoftBank and Honda were able to predict such behaviors and issue timely alerts to drivers, mitigating the risk of accidents. The trials revealed challenges in synchronizing data streams: roadside sensor inputs and vehicle telemetry arrived with differing delays, complicating accurate reconstruction of traffic conditions in the digital environment. To address this, the team implemented a system that compensates for timestamp mismatches and latency variations.
From an engineering perspective, the integration of heterogeneous data sources—vehicle-borne sensors, roadside infrastructure, and cellular networks—demonstrates the complexity of real-world V2X deployments. Achieving low-latency, high-reliability communication is essential for safety-critical applications, particularly when predicting and responding to rapid changes in vehicle trajectories.
The project underscores the importance of scaling connected vehicle penetration and expanding roadside sensor coverage. With broader deployment, such systems could enable comprehensive information sharing across entire road networks, enhancing situational awareness for all drivers and reducing the likelihood of collisions. The collaboration between telecommunications and automotive engineering teams illustrates how cross-industry partnerships can accelerate the development of advanced mobility solutions.