Automated Transport Vehicle for Visually Impaired Assist at Indoor Transportation

Abstract

This paper presents a low-cost Automated Transport Vehicle (ATV) designed to assist individuals with visual or musculoskeletal disabilities in indoor navigation. The platform is based on a commercially available electric scooter (Optimus Up) modified with a Programmable Logic Controller (JZ20-T18 PLC), two DRV5032AJ linear Hall-effect sensors for magnetic-path following, Bluetooth Low Energy (BLE) beacons (HC-08) for coarse indoor localization, and a voice-command interface (SnowBoy, with a planned replacement noted under System Architecture) handled by a Raspberry Pi. The vehicle follows electromagnetic routes selectively activated by a relay matrix (Navigation Map), enabling destination-based routing across a five-room environment. The login control algorithm operates on a differential sensor comparison scheme: lateral imbalance between left and right sensors triggers corrective steering via electromagnetically actuated wheel control. Performance is evaluated in a controlled environment, yielding a mean lateral error of 1.52 cm, an RMSE of 2.25 cm, a 100% path-completion rate, an average control-system response time of 0.558 s, and a fail-safe stopping distance of 7.37 cm. The BLE localization subsystem was validated independently in a prior study but has not been evaluated as part of the complete ATV platform. A fuzzy-logic module estimates route-deviation risk under varying operating conditions.