The smart backseat driver you might welcome
According to the influential Task-Capability Interface model, road collisions increase when drivers are not in control, meaning their capacity doesn’t match ‘task’ demands. Varying during a trip, tasks are determined by situational factors such as speed, road condition or weather; with the driver’s capacity influenced by skills and expertise, and human factors such as being tired. “While drivers adapt their driving to meet tasks – such as avoiding mobile phones during difficult driving – tasks and abilities can be easily under- or overestimated,” explains Tom Brijs, project coordinator of the EU-funded i-DREAMS project. i-DREAMS studied behaviour in natural driving conditions to validate aspects of the Task-Capability Interface model and pinpoint areas of potential intervention. The team developed and validated a concept, the ‘Safety Tolerance Zone’ (STZ), of three phases – normal driving, danger, avoidable accident – then designed and tested smart safety interventions (in-vehicle and post-trip).
From warnings to rewards
The i-DREAMS system consists of an integrated suite of sensors that can be retrofitted to vehicles. If breaches to safe driving boundaries are detected, real-time in-vehicle visual and/or audio warnings alert the driver – becoming more intrusive with increasing risk. A video of the situation is recorded and sent to the cloud for later evaluation. Tailored warnings were developed for risk indicators such as: handheld mobile phone use, fatigue, harsh acceleration, deceleration or steering, tailgating, lane change without indicators and speeding. Post-trip, the i-DREAMS app uses gamification techniques to coach drivers, using trip records, including risk events and associated videos, to calculate a safety score for each indicator and the overall trip. Based on performance over several trips, drivers receive advice and can sign up for personalised goals. Once reached, participants receive a reward badge (bronze, silver, gold, platinum) and earn ‘safety credits’, exchangeable in the app’s digital shop.
The rubber hits the road
The i-DREAMS system was first tested in different driving simulators (car, truck, bus, train), before piloting vehicles in real road conditions. Finally, large-scale on-road field trials were held in five countries (Belgium, Germany, Greece, Portugal, United Kingdom) and in three vehicle types (car, bus, truck). In these trials, drivers were exposed to increasingly intense interventions, from monitoring only, to activation of real-time warnings and tips, all the way to the system’s gamification functionalities. All 650 drivers (250 in simulators, 400 in field trials) completed each stage over a four and a half month period, with over 3.4 million miles driven, taking over 300 000 hours and equating to 120 000 trips. Afterwards, the data was analysed to determine how driver behaviour changed over the course of the interventions. “We did find that risky events decreased as drivers progressed through the intervention levels, with the post-trip coaching clearly boosting safer driving. However, differences remain between drivers and countries. Also, the interventions had less impact on professional drivers,” says Brijs from Hasselt University, the project host.
The road ahead
The i-DREAMS system aims to instil a safer driving culture. To help embed this, an i-DREAMS web dashboard was created for transport companies, including a mentor/coach assigned to drivers. The project consortium has already reached an agreement with its industrial partners to commercialise the technology and is now negotiating the first contracts with transport companies. “I’m proud that this technically, legally and ethically complex project, can be commercially launched almost immediately, helping to create safer roads,” concludes Brijs. The team is now exploring how its anonymised raw and processed data could help teaching facilities and policymakers better understand driving behaviour, with an initial data extract available on Zenodo.
Keywords
i-DREAMS, collision, driver, safe, road, driving, vehicle
