| Summary: |
Objectives/Activities: The impairing effects from sleepiness and distraction are a major contributor to road crashes. Industry has responded with investment in driver monitoring tools aimed at mitigating the risk. That is the case of the Healthy Road (HD) device developed by the HealthyRoad - Biometric Systems, Lda©, a Portuguese start-up company. The HD system uses an infra-red emitter and sensor mounted on a spectacle frame to continuously measure eye blink velocity, from which levels of drowsiness are derived. Additionally, distraction is also monitored based on the face position and eyes direction. A huge quantity of data (big data) has been recorded regarding alerts (number, time and location of each emitted alert) as well as driver's and vehicle's characteristics and journey. This big data is valuable to several scientific researches, offering an opportunity to develop novel studies based on real-time data.
In this context, the project SIESTA aimed to analyze the collected data in order to assess drowsiness and distraction patterns taking into account the driver's journey's characteristics. This latter includes the local where the alerts were emitted, which in time gives the information about the road (type, infrastructure and environment), the length of the journey. Three main tasks were developed: T1 Data treatment and analysis; T2 Identification of relevant patterns; T3 Writing Final Report and research papers.
Results/Impact: Two main goals were addressed, namely identification and analyses of risk factors and drivers' profiles, based on two distinct statistical techniques. Regarding risk factors analysis, the results suggest for instance that increasing the driving continuous time of professional drivers, the number of alerts of both types increase too. If the driver stops during the journey, the number of alerts, either distraction and drowsiness, decreases. The duration time of the breaks revealed also an effect on  |
Summary
Objectives/Activities: The impairing effects from sleepiness and distraction are a major contributor to road crashes. Industry has responded with investment in driver monitoring tools aimed at mitigating the risk. That is the case of the Healthy Road (HD) device developed by the HealthyRoad - Biometric Systems, Lda©, a Portuguese start-up company. The HD system uses an infra-red emitter and sensor mounted on a spectacle frame to continuously measure eye blink velocity, from which levels of drowsiness are derived. Additionally, distraction is also monitored based on the face position and eyes direction. A huge quantity of data (big data) has been recorded regarding alerts (number, time and location of each emitted alert) as well as driver's and vehicle's characteristics and journey. This big data is valuable to several scientific researches, offering an opportunity to develop novel studies based on real-time data.
In this context, the project SIESTA aimed to analyze the collected data in order to assess drowsiness and distraction patterns taking into account the driver's journey's characteristics. This latter includes the local where the alerts were emitted, which in time gives the information about the road (type, infrastructure and environment), the length of the journey. Three main tasks were developed: T1 Data treatment and analysis; T2 Identification of relevant patterns; T3 Writing Final Report and research papers.
Results/Impact: Two main goals were addressed, namely identification and analyses of risk factors and drivers' profiles, based on two distinct statistical techniques. Regarding risk factors analysis, the results suggest for instance that increasing the driving continuous time of professional drivers, the number of alerts of both types increase too. If the driver stops during the journey, the number of alerts, either distraction and drowsiness, decreases. The duration time of the breaks revealed also an effect on the alert frequency. The results for non-professional drivers showed the same kind of trends, despite same differences. The results of the clustering analyses revealed at least three clusters of drivers that can be distinguish in terms of journey characteristics and alert frequency, allowing a clear classification. These findings were compared with other studies whenever was possible. Overall, the project showed that data from emerged technology has potential to be explored to road safety studies. |