Biometric Interface for Driver's Stress Detection and Awareness

One of the factors for stress reduction among vehicle drivers is to be aware that stress is present. This project presents a biometric interface for stress detection in drivers, built with open source sensors and hardware. In two series of experiments, we induce stress in test subjects by making them drive progressively difficult scenarios in a simulator. Using the C4.5 classification algorithm, we classified the subjects' biometric data in order to determine whether the subject was stressed or not. In another series of experiments, we tested the efficacy of two driver feedback systems, a haptic one and a visual one. Identifying a stressful situation allows real-time feedback to drivers, so they can be aware of their stressed state, thus being able to take corrective actions on time, and avoid behavior leading to an accident.