Abstract
Safety in road tunnels are of utmost importance for the public notion of safety within the road system. In recent years, there has been significant progress in multiple areas of artificial intelligence, sensor fusion and communication technologies. Together with increase in computing power, this has enabled processing capabilities and aggregation of large amount of data from heterogenous sources. This allows for more intelligent decision-making in real-time in presence of risk in a dynamic environment provided by a decision support system. Previous work in this direction do not actively combine risk-awareness, real-timeness, and artificial-intelligence in a dynamic operational environment of a tunnel in operation for decision-making through considering the capabilities that recent technological advancements enable. To address this gap between decision-support systems and state-of-the-art technologies, this paper proposes RiskTUN, a general framework for developing risk-aware decision support systems for the safety of tunnels in operation. RiskTUN architecture allows for integration of various sources of data in a heterogenous environment where various stakeholders (e.g., road users, emergency responders, traffic centers, etc) can be both contributors or the users of the decision support system. There are major opportunities associated with taking better advantage of available data, but challenges are also identified and discussed. System implementations made based on RiskTUN framework are expected to better adapt to the user needs within the area of tunnel safety as technologies evolve.