Multipurpose sensing platform for improved road safety

• Autor(a) principal: Carvalhosa, Miguel Filipe Pereira de Freitas
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/34076
• topic: Roadside sensors; Road safetys; Weather monitoring; Collision monitoring; Traffic monitoring; Low-power; Energy harvesting; Embedded systemss; Internet of things; LoRaWAN; MQTT; InfluxDB; Grafana

With the increasing number of vehicles in the European roads and the emergence of autonomous vehicles and their coexistence in the roads with legacy vehicles, road safety is still a primary concern. This dissertation describes the design and implementation of a multipurpose roadside sensing platform for improved road safety. It is capable of monitoring weather conditions in highways with large coverage and high granularity, allowing road operators to predict dangerous driving conditions and alert drivers in real-time. Moreover, the sensing platform is equipped with an accelerometer to detect impacts on the highways’ guardrails and transmit emergency messages, improving the response of emergency services and allowing for proper maintenance to ensure the integrity of the guardrails. It also features a microphone and a lightweight vehicle detection algorithm based on sound, to monitor the traffic intensity. The devised solution is scalable and low-cost, due to the simple design using commonly available components. It features low communication latency and low power consumption, due to the use of a LoRaWAN network and low-power components, as well as the design choices to keep the device in sleep mode when not performing any actions. The sensors require low maintenance and are battery powered with solar energy harvesting, meaning that they can be placed in strategic and dangerous areas of highways, where there are no electricity sources, and be left operating standalone without the need of on-site human interaction. This solution is enabled due to the latest developments in embedded systems and Internet of Things technologies, such as LoRaWAN, MQTT, InfluxDB, and Grafana. This dissertation presents in detail the architecture of the different modules that integrate the system, as well as their implementation. The behavior of the system was verified with tests of the individual components and real-world field tests, such as the validation of the communication latency, vehicle detector, battery autonomy and solar charging, and weather and collision monitoring.