Driver Alertness Comparison Using BCI Data between the Voice-Based Arithmetic System and Traditional Audio and Visual Alerts

About 36,550 people lost their lives in vehicular accidents caused by unresponsiveness and drowsiness in 2018 [1]. Despite advancements in driver-assisted technologies and some automobile industries’ attempts to design technologies to alert drivers in a monotonous driving environment, there is a lack of relevant technology that notably improves driver alertness and engages drivers. We develop a Voice-Based Arithmetic System (VBAS) to engage drivers using simple arithmetic questions. During experiments, participants drove in a simulated driving environment to compare their drowsiness using the proposed VBAS versus traditional audio and visual alerts. Variation in brain signals was captured using a Brain-Computer Interface device to detect drowsiness levels. We analyzed the video recordings of sessions to compute the blink and yawn rate of participants. Participant sleep data from the night before the experiment was recorded using a Fitbit Alta. This research highlights the potential of similar technologies that can improve the alertness of drivers. According to both alpha and theta brain signals, participants were less drowsy after using the VBAS than traditional audio and visual alerts. Mitigating the drivers’ drowsiness is a critical step in decreasing the number of associated crashes and fatalities. The Voice-Based Arithmetic System has the potential to engage drowsy drivers longer than audio and visual alerts and help reduce the likelihood of drowsy-driving related crashes.