This thesis describes the development of the backend of a system for monitoring residential environments through a network of sensors installed in households. The core of the work is the implementation of a Server Model Context Protocol (MCP), which leverages an innovative protocol to interface the system’s functionalities with large language models (LLMs). This approach enhances the analysis of the data collected by the sensors with artificial intelligence tools, paving the way for advanced monitoring and support solutions for managing domestic environments.
La presente tesi descrive lo sviluppo del backend di un sistema per il monitoraggio di ambienti abitativi attraverso una rete di sensori installati nelle abitazioni. L’elemento centrale del lavoro è l’implementazione di un Server Model Context Protocol (MCP), che sfrutta un protocollo innovativo che consente di interfacciare le funzionalità del sistema con modelli di intelligenza artificiale (LLM). Questo approccio permette di arricchire l’analisi dei dati raccolti dai sensori con strumenti di intelligenza artificiale, aprendo la strada a soluzioni avanzate di monitoraggio e supporto per la gestione degli ambienti domestici.
Server Model Context Protocol per il monitoraggio di ambienti abitativi
MILAN, RICCARDO
2024/2025
Abstract
This thesis describes the development of the backend of a system for monitoring residential environments through a network of sensors installed in households. The core of the work is the implementation of a Server Model Context Protocol (MCP), which leverages an innovative protocol to interface the system’s functionalities with large language models (LLMs). This approach enhances the analysis of the data collected by the sensors with artificial intelligence tools, paving the way for advanced monitoring and support solutions for managing domestic environments.| File | Dimensione | Formato | |
|---|---|---|---|
|
Milan_Riccardo.pdf
Accesso riservato
Dimensione
2.36 MB
Formato
Adobe PDF
|
2.36 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/102064