FMWare: IoT-Based Fleet Management System

Abstract

The Los Angeles County Department of Parks and Recreation oversees over 70,000 acres and manages 182 parks across the County. The Department uses over 600 vehicles to cover this vast area. Currently, the Department uses a paper-based system to manage these vehicles. However, the paper-based system cannot effectively schedule a regular maintenance. This can result in safety hazards for drivers as well as poor management of the County resources. In addition, detecting unauthorized use of vehicles is very difficult. Current proprietary industry solutions are too costly for the Department; yet the system in use by the Department is unable to serve the County’s needs. In this paper we propose FMWare, a low cost, fully automated IoT-based fleet management solution. The Edge portion of FMWare is composed of a Raspberry Pi 3 B+ , a vehicle on-board diagnostics (OBDs) adapter, a GPS module, an RFID reader, and a custom power supply unit (PSU). To use a vehicle, an employee must scan their I.D. badge to the RFID reader. This ensures that the driver is identified for each trip. Unauthorized use of County vehicles is detected as well. When a vehicle is moving, FMWare collects the vehicle’s diagnostics and GPS data. The collected data is saved in the on-board memory of the Raspberry Pi. Upon returning to a base station, the collected data from the vehicle is uploaded to the County Cloud server through secure wireless gateways available at the County parking lots. We use Apache NiFi and MiNiF, open-source data transmission control software, to upload the collected data. Given the volume of data, the overhead of a traditional RESTFul Web Service would overwhelm the server. Experiments demonstrate the feasibility of the FMWare data collection method and the overall design of the Edge data collection platform.