Exploratory MCDA for handling deep uncertainties: The case of intelligent speed adaptation implementation

Sometimes experts, decisionmakers, and analysts are confronted with policy problems that involve deep uncertainty. Such policy problems occur when (1) the future is not known well enough to predict future changes to the system, (2) there is not enough knowledge regarding the appropriate model to use to estimate the outcomes, and/or (3) there is not enough knowledge regarding the weights stakeholders currently assign to the various criteria or will assign in the future. This paper presents an MCDA approach developed to deal with conditions of deep uncertainty, which is called Exploratory Multi-Criteria Decision Analysis (EMCDA). EMCDA is based on exploratory modelling, which is a modelling approach that allows policy analysts to explore multiple hypotheses about the future world (using different consequence models, different scenarios, and different weights). An example of a policy problem that can benefit from this methodology is decision making on innovations for improving traffic safety. In order to improve traffic safety, much is expected from Intelligent Speed Adaptation (ISA), an in-vehicle system that supports the driver in keeping an appropriate speed. However, different MCDA studies on ISA give different results in terms of the estimates of real-world safety benefits of ISA and the willingness of stakeholders (e.g. the automotive industry) to supply ISA. The application of EMCDA to the implementation of ISA shows that it is possible to perform an MCDA in situations of deep uncertainty. A full analysis taking into account the complete uncertainty space shows that the best policy is to make mandatory an ISA system for young drivers (less than 24 years of age) that restricts them from driving faster than the speed limit. Based on different assumptions, the analysis also shows that ISA policies should not target older drivers.