Crowdsensing-Based Road Condition Monitoring Service: An Assessment of Its Managerial Implications to Road Authorities

Abstract

The ubiquity of smart devices in vehicles, such as smartphones allows for a crowdsensing-based information gathering of the vehicle’s environment. For example, accelerometers can reveal insights into road condition. From a road authorities’ perspective, knowing the road condition is essential for scheduling maintenance actions in an efficient and sustainable manner. In Germany, expensive laser-based road inspections are scheduled every four years. In future, they could be extended or completely replaced with a crowd-based monitoring service. This paper determines whether the lower accuracy of crowdsensing-based measurements is redeemed by its potential of near-real time data updates. Partially observable Markov decision processes are applied for determining maintenance policies that minimize roads’ life-cycle costs. Our results show that substituting laser-based road condition inspections by a crowdsensing-based monitoring service can decrease total costs by 5.9 % while an approach, which combines both monitoring approaches, reduces the costs by 6.98 %.