i-DREAMS: D2.2 Technologies for safety interventions and assessment of their effectiveness

Abstract

This deliverable aims to compare and contrast the available technologies for safety interventions both in real-time (i.e. during a driving session) as well as post-trip. These safety interventions will be a significant module within the i-DREAMS project, as they will inform or warn the driver with regards to the real-time safety level, and will provide a gamified coaching platform to enhance longtime driving performance or skills. The findings of this review, form the results of Task 2.2, within WP2 of the i-DREAMS project. Initially, the definition of a safety intervention is provided based on the literature, and the importance of a correct intervention strategy is highlighted. Furthermore, theoretical principles related to acceptance and performance are analyzed, and recommendations for designing the i-DREAMS intervention strategy are given. It was shown that acceptance should be pursued closely with the operators participating in the simulator and naturalistic driving experiment using survey and observational techniques, and that training could enhance acceptance of technologies. With regards to performance, a multi-stage multimodal approach on feedback was found to be an advantageous option. Real-time and post-trip intervention technologies were assessed based on the criteria of acceptance and effectiveness distinctively for each mode in the i-DREAMS project (i.e. car, truck, bus and train). For each of the modes, commercially available technologies, as well as interventions tested within the academic literature, were overviewed and assessed. At the end of each corresponding mode-specific review, recommendations on the intervention technologies per mode are given, and cross-modal considerations and differences between professional and non-professional drivers are discussed. More details on legislation, regulations, as well as targeted theories on interventions for professional drivers are discussed in the sections for truck and bus operators. The in-depth literature review revealed that eye-tracking and similar visual sensors are deemed the most efficient for monitoring driver state and provided interventions in real-time with auditory or visual messages. Furthermore, if such devices are of small size or are connected with a smartphone application they can easily be modified and transferred, but there is almost a concern on practical implementations of these tools. However, the state-of-the-art in such intervention technologies are application specific and do not provide an holistic intervention approach to driver performance degradation. Gamification and an appropriate reward or penalty system were proven to be the most effective strategy for post-trip interventions, as it was found that they keep drivers motivated in order to enhance their driving skills. With regards to professional drivers, coaches can educate them on an optimal driving behaviour, after receiving an overview of drivers' performances. Nevertheless, it should be mentioned that in trucks and buses, interventions are usually part of a broader framework (i.e. including driver coaching and management commitment) and the effects of such interventions cannot be taken into account in isolation for accomplishing a sufficienrt safety culture changeIt is worth mentioning that there was not found a complete post-trip intervention solution massively developed or tested in real-world environments. For each of the considered transportation modes suggestions were made with regards to monitoring technologies and commercial solutions. The provided recommendations and considerations for the safety interventions to be included in the i-DREAMS platform, need to also be considered for the mathematical formulation of the Safety Tolerance Zone. In addition, a compromise needs to be found between the estimation of the safety level, the triggering and information provided in real-time, and the analytics or coaching provided after the end of driving trips.