Experimental evaluation of a traffic warning system based on accurate driver condition assessment and 5G connectivity

Abstract

Reliable detection and sharing of information about fatigued or otherwise impaired drivers can provide valuable extra information to improve cooperative road traffic safety services. With such information, connected manually driven or automated vehicles in the area can proactively take precautions and prepare for possible risks caused by a fatigued drivers. In order to provide accurate assessment of the driver condition and efficient distribution for the related warnings, the proposed human tachograph service concept combines ubiquitous wearables-based driver monitoring with 5G connectivity. The combination of the real-time driver biosignals measured while driving and the historical data related to driver’s sleep and physical activity outside the vehicle enables the driver condition to be assessed more accurately than with currently used on-board systems. Based on the driver condition analysis, the 5G-based traffic warning system triggers warning messages towards other road users. The paper also presents the trial setup used to evaluate the performance of end-to-end service as well as the 5G network on top of which the service is deployed. Based on the results, the initial 5G deployments can already achieve clearly better average latency than LTE-based deployments but the reliability should yet be improved for road safety applications.