A carregar...
Publicação
LoRaWAN Based Tracking System for Sport Events on Mountain and Forest Areas
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 11.69 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Terrenos inóspitos, como as florestas e as montanhas, são populares para
praticar desportos ao ar livre como o trail, BTT, motocross e caminhadas, que,
apesar de emocionantes, têm riscos inerentes. Os atletas enfrentam desafios
como o esforço físico intenso, condições climatéricas extremas, desidratação,
percursos complexos e a possibilidade de acidentes, que podem provocar
lesões ou desorientação. Uma resposta rápida nestes cenários é crucial para
orientar os atletas perdidos, tratar de lesões ou mesmo salvar vidas. Nestes
terrenos, normalmente não há conectividade Wi-Fi ou rede móvel, o que
dificulta a localização dos atletas que precisam de ajuda e a rápida prestação
de assistência.
A implementação de um sistema capaz de localizar continuamente os atletas
durante as competições aumentaria a segurança e garantiria a exatidão dos
percursos. Esta tese implementa a primeira geração deste sistema de
localização. O sistema é composto por uma rede utilizando LoRaWAN, que foi
criada através do desenvolvimento de um gateway e de nodes com
capacidades de GPS (que estariam agarrados a cada atleta). Depois de
adicionar um sistema de backend, o atleta (node) foi possível de rastrear no
campo através de um painel de controlo. Foram realizados vários testes no
terreno, que demonstraram o sucesso das operações do sistema, bem como a
sua fiabilidade e capacidade de recolher dados sobre a forma como cada
ambiente e configuração altera os resultados. Estes dados recolhidos em
massa foram de seguida utilizados para desenvolver modelos de Machine
Learning que conseguem prever o comportamento das transmissões no terreno
com elevada precisão, com o objetivo de melhorar o desempenho geral do
sistema.
Rough terrains like forests and mountains are popular for outdoor sports like trail running, cycling, motorcycle riding, and hiking, which, while exciting, come with inherent risks. Athletes face challenges such as intense physical exertion, extreme weather conditions, dehydration, complex routes, and the potential for accidents, leading to injuries or disorientation. Quick response in such scenarios is crucial to guide lost athletes, address injuries, or even save lives. These terrains usually lack Wi-Fi or cellular connectivity, making it difficult to locate the athletes in need of help and quickly provide assistance. Implementing a system capable of continuously tracking athletes during competitions would enhance safety and ensure the routes’ accuracy. This thesis implements the first generation of this tracking system. The system is composed of a network using LoRaWAN, which was created by developing a gateway and nodes with GPS capabilities (that would be attached to each athlete). After adding a backend system, the athlete (node) can be tracked in the terrain by monitoring a dashboard. Several field tests were performed and demonstrated the successful system operations, as well as its reliability and capability to gather data on howeach environment and setting alters the results. This massive collected data was then used to develop Machine Learning models that can predict the behavior of the transmissions in the field with high accuracy, with the goal of improving the system’s overall performance.
Rough terrains like forests and mountains are popular for outdoor sports like trail running, cycling, motorcycle riding, and hiking, which, while exciting, come with inherent risks. Athletes face challenges such as intense physical exertion, extreme weather conditions, dehydration, complex routes, and the potential for accidents, leading to injuries or disorientation. Quick response in such scenarios is crucial to guide lost athletes, address injuries, or even save lives. These terrains usually lack Wi-Fi or cellular connectivity, making it difficult to locate the athletes in need of help and quickly provide assistance. Implementing a system capable of continuously tracking athletes during competitions would enhance safety and ensure the routes’ accuracy. This thesis implements the first generation of this tracking system. The system is composed of a network using LoRaWAN, which was created by developing a gateway and nodes with GPS capabilities (that would be attached to each athlete). After adding a backend system, the athlete (node) can be tracked in the terrain by monitoring a dashboard. Several field tests were performed and demonstrated the successful system operations, as well as its reliability and capability to gather data on howeach environment and setting alters the results. This massive collected data was then used to develop Machine Learning models that can predict the behavior of the transmissions in the field with high accuracy, with the goal of improving the system’s overall performance.