A IoT-Enabled Electric Scooter Rental System with Docking Stations RAFEEQ

Abstract

rban mobility in Algerian cities suffers from a structural void in the one-to-three-kilometre trip range: public buses are too rigid, taxis too expensive, private cars too inefficient, and walking too constrained by climate and infrastructure. This thesis presents the design, im- plementation, and validation of Rafeeq, a smart electric scooter rental platform conceived to address this First and Last Mile problem in the Algerian institutional, economic, and cultural context. The platform integrates four interoperable subsystems. A connected fleet of Xiaomi M365 electric scooters is instrumented with a Raspberry Pi 3 Model B+ onboard computer paired with a SIM800L 2G GPRS modem for cellular uplink, and communicates with the scooter’s electronic speed controller through the Bluetooth Low Energy Nordic UART ser- vice. A network of physical docking stations equipped with ESP32 microcontrollers drives 12 V electromechanical Solenoid locks through TIP122 Darlington power stages, with full electrical protection including flyback diodes and LM7805 regulated rails. A cloud backend implemented in FastAPI and deployed on Render exposes thirty-two REST endpoints and maintains real-time WebSocket gateways with both the IoT endpoints and the mobile clients; user data is persisted in a PostgreSQL 16 database with strict ACID guarantees on all wallet operations. A cross-platform Flutter mobile application covers sixteen user screens across the full rental journey, with full right-to-left Arabic localisation, French, and English support. The system was validated through a six-family experimental campaign covering elec- tronic bench tests, BLE communication, mechanical lock characterisation, end-to-end latency measurement, backend load tests, and mobile user trials. The prototype achieves an end-to- end unlock latency of 2.83 seconds against a 3-second target, a Solenoid holding force of 73 N against a 50-N target, an onboarding completion time of 78 seconds against a 90-second target, and a station continuous power consumption of 84 W against a 120-W ceiling. The validation demonstrates that an integrated, locally developed smart-mobility platform can technically address the First and Last Mile problem in the Algerian context, and positions Rafeeq as a credible candidate for a subsequent pilot deployment at the University Mohamed El Bachir El Ibrahimi campus.