EdgeAI - Bringing AI to Smart Devices (Drowsiness Detection System)
Year: 2026
Type: article
Abstract: Driver drowsiness can reduce attention and reaction capability, making driver-state monitoring an important road-safety research area. This paper presents a proposed design for a low-complexity embedded driver-warning system using an infrared (IR) eye-related sensor, a controller, and visual and audible alerts. It outlines the problem context, relevant sensing approaches, functional architecture, general decision flow, and a controlled evaluation protocol to be used after implementation. The paper also distinguishes rule-based embedded processing from Edge AI: a system should be described as Edge AI only when an AI model is deployed and executed locally on or near the data source. The present manuscript is a design-stage study; the physical model, exact circuit, source-code behaviour, and controlled measurements are not presented as completed or verified. Accordingly, no accuracy, latency, power-consumption, or reliability results are claimed. The proposed evaluation will examine sensor response, alert behaviour, false warnings, missed detections, recovery, and repeatability in a stationary controlled environment. The work provides a transparent design framework that can be refined after the hardware and software are implemented and tested. The proposed device is intended as an educational prototype and not as a certified automotive safety system.
Authors: Srushti Joshi, Rohini Chipade, Suvarna Amale
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7225001347