After 7 days of machine learning and calculation through the cloud, the monitoring system can run error detection and diagnosis, and proactively notify designated maintenance personnel via mobile phone or email.

There is no need for additional hardware but to import data from on-site equipment such as insolation meters and MPPT (solar controllers), and no restriction on the brand of the inverter.  The expected power generation curve of each specific site can be established within 7 days, and then can be matched with the real-time generation for monitoring and error diagnosis. 

This technology has been recognized internationally at the EU Photovoltaic Solar Energy Conference(EUPVSEC), the largest PV solar technical research forum globally, in 2019 & 2020, by inviting our representatives Kun-Hong Chen & Mao-Yi Chang to have oral presentations and ranking the research No.23 in 900+ papers worldwide. In 2021, we were also invited by Intersolar, the world's leading exhibition for the solar industry, to present the research online during the Covid period.

By applying the technology we developed, the maintenance operation efficiency and power generation were steadily improved, further reducing the maintenance carbon footprint by 40%.