On-board unit to connect personal mobility vehicles to the IoT

Abstract

Abstract The basis of smart city services is urban sensing, i.e., the collection of data from sensors deployed in urban settings. However, it has been recently when mobile objects have started to be monitored to feed smart applications within the smart city ecosystems. Most of the advances in this area focus on gathering data from public transportation services such as buses or trams, which use to employ regular cellular or WiFi-based networks and state-of-the-art TCP/IP protocols. Although this covers a basic on-board equipment to report position and other sensor data, there is not a direct integration of Internet of Things (IoT) capabilities in urban transportation and, in general, in moving “things”. Communication technologies used in regular wireless sensor networks such as ZigBee or Bluetooth are not appropriate for high mobility patterns. The Internet of Vehicles (IoV) paradigm aims at filling these gaps, exploiting the synergy between IoT and vehicular networks. It is an emerging area in which extra efforts should be paid to cover constrained capabilities of end devices installed on personal mobility devices, e.g., scooters, segways or bicycles. The solution proposed here deals with the connectivity of moving objects by developing an on-board unit for energy-constrained vehicles, integrating communication capabilities using the Low-Power Wide Area Network (LPWAN) technologies Long Range Wide-area Network (LoRaWAN) and Narrow-Band IoT (NB-IoT) to efficiently collect data. A set of sensors are included to target services in the areas of urban mobility, enhanced environmental perception and speed-up of city operations. The on-board unit has been evaluated through real-life experiments to study especially its communication capabilities with NB-IoT, obtaining good performances in terms of packet delivery ratio and network delay under non-ideal coverage settings. Results show that NB-IoT presents better performances than LoRaWAN, as compared to our previous experiences, with negligible packet loss rates and two-way delays under 100 ms with Internet nodes, at the expense of needing a proper operator network deployment and consuming extra energy due to NB-IoT protocol complexity.