Solar cell performance is often limited by defects such as cracks, which are visible in luminescence images. Identifying these defects and understanding their impact are critical for both manufacturing and long term reliability. In this project, we will use explainable artificial intelligence (AI), a set of techniques that help us understand how AI models work, to study these images. In this context, explainable AI will be used to highlight the regions of each image most relevant to detecting and assessing defects. There are many models and explainable AI methods available, and students can choose which approach they want to explore.
The project will be carried out in a collaborative research environment. Students will mainly work with one or two academic researchers and be part of a research team that includes undergraduate and postgraduate students.
The aim of this project is to investigate how explainable AI can support a better understanding of solar cell performance through defect analysis. An interest in Python programming is encouraged. Even a basic familiarity will help you get the most out of the project, and we will support you in developing your skills further.
School
Photovoltaic and Renewable Energy Engineering
Research Area
Explainable AI | Deep learning | Solar cells | Luminescence imaging | Computer vision
Suitable for recognition of Work Integrated Learning (industrial training)?
No
- Research environment
- Expected outcomes
- Supervisory team
- Reference material/links
You will work with the ACDC Research Group, a lively team of more than 20 researchers dedicated to solar energy research. The group has a friendly and supportive environment. You'll benefit from close mentoring, regular meetings with supervisors, and opportunities to share your work with other team members. In addition to your research, you'll also be able to take part in a variety of social activities with the team.
- Students will gain experience with image based machine learning models and explainable AI.
- They will apply these methods to solar cell image data and present insights into how transparent models can support quality assessment.
- Relevant work can be found on the website of the ACDC Research Group: https://www.acdc-pv-unsw.com/publications