This research project aims to develop a novel platform for the real-time acquisition of IV characteristics of photovoltaic (PV) modules, leveraging low-cost devices with Long Range Wide Area Network (LoRaWAN) connectivity for long-range data transmission. The project aligns with the mission of the Levi Cases Centre, focusing on renewable energy research, particularly in energy production, transformation, distribution, and usage. By accurately tracking the electrical/thermal/optical characteristics of PV modules, the project seeks to gather fundamental operating data from module, to i) optimize renewable energy utilization, ii) detect failure and different failure modes in modules, iii) provide data set and mathematical models to train artificial intelligence algorithms with the final objective to improve photovoltaic systems efficiency, reliability, predictive maintenance and production prediction.

The aim of the project is to develop a small, low cost and low power consumption measurement system to be coupled with PV modules, thus forming a distributed sensor network able to measure parameters of the module which cannot be measured at a system level.

First, the structure of the Electronic Sensing Unit (ESU) will be defined and implemented, then, the communication protocol will be developed, in particular LoRaWAN will be used since this type of transmitters are already integrated into many low-cost microcontrollers. The total information related to the module characteristic can be transmitted with just two frames.

Finally, a generative adversarial algorithm to train the AI predictor with maliciously generated data will be developed. The objective of this training methodology is to develop an optimization framework that cannot be exploited by an attacker by simply injecting fake data in the power network.

The platform's prototype will be thoroughly tested in real-world scenarios, ensuring its readiness for large-scale deployment.

Additionally, this project emphasizes the utilization of acquired data to not only optimize production processes but also to explore broader implications, including cost-benefit analyses and potential social impacts. The data that will be made available by a large adoption of the system developed in this project can be employed by AI tools and algorithms, thus allowing the extraction of valuable insights from the collected data, contributing to advancements in renewable energy research and beyond.