Characterisation, defects & machine learning

Making the most energy-conversion efficient solar cell requires careful inspection of the materials used to make them and of the fabrication processes along the entire production chain. Identifying the causes of efficiency losses in solar cells, and developing ways to mitigate or eliminate them, is an area of our research focus. 

We also investigate methods to assess and improve the reliability of photovoltaic cells, modules, and systems. A special focus of our group is photoluminescence imaging -- a technique that was developed at UNSW. In the last decade, photoluminescence imaging has become a standard characterisation process in the photovoltaic industry, utilised by almost every silicon solar cell manufacturer in the world.

Our group has pioneered the use of machine learning and deep learning for photovoltaic applications in the characterisation of defects in photovoltaic devices. We are using our extensive expertise in interpreting correlations between various metrics and solar cell performance to establish a new research area that combines photovoltaics and computer science.

Our research spans a wide variety of new applications for photovoltaics, such agri-photovoltaics (Agro-PV), building-integrated photovoltaics (BIPV), and Internet of Things (IoT).

Recently, our students have won more than ten Best Student Awards at leading international conferences, highlighting our position as a world-leading photovoltaic research group. 

For more information visit the ACDC Research Group website.