Photovoltaics (PV) is now the lowest‑cost source of new electricity generation, but sustained gains in manufacturing yield rely on high‑throughput, quantitative characterisation. As cell architectures adopt advanced metallisation (e.g., multi‑busbar, shingled/0‑BB, back‑contact and tandem stacks), reliable contacting for conventional tests becomes error‑prone and slow. Our ACDC (Artificial intelligence, Characterisation, Defects, and Contacts) research group at SPREE has therefore developed contactless, photoluminescence workflows to extract key electrical parameters. A core challenge is the accurate determination of short circuit current (Jsc), the maximum current the cell can produce, which is directly related to its efficiency. Optimising speed and accuracy is vital, especially for high‑throughput manufacturing lines as faster measurements will be less accurate, especially for next‑generation multijunction cells. This ToR project targets this gap by utilising fast, accurate, and generalised machine learning (ML) models to extract Jsc from lower resolution measurements.
You will work with an experienced team in PV characterisation methods, tool development, and ML applications. Within this team, you will explore advanced characterisation methods and equipment, apply electrical and optical modelling to silicon and next‑generation devices, and train/validate ML models to rapidly and accurately estimate Jsc across diverse cell designs. The final outcome is an ML‑assisted pipeline paired with our contactless Jsc equipment, advancing the development and commercial potential of vital PV characterisation tools.
Your supervisory team, Dr Zubair Abdullah‑Vetter, Dr Arthur Julien, and Prof. Ziv Hameiri, will play pivotal roles to ensure a successful ToR experience and outcomes. Supervisory team profile:
- Dr Zubair Abdullah‑Vetter, Postdoctoral Fellow (ACDC, SPREE): specialising in ML for advanced PV characterisation, including image analysis and quantum efficiency modelling.
- Dr Arthur Julien, Research Fellow (ACDC, SPREE): specialising in experimental/theoretical PV physics focused on tandem and perovskite devices, advanced characterisation, and modelling.
- Professor Ziv Hameiri, Group Lead, ACDC (SPREE, UNSW): leading ACDCs programs in contactless characterisation, machine‑learning for PV, and defect/interface investigations across silicon and tandem devices.
School
Photovoltaic and Renewable Energy Engineering
Research Area
Machine learning | Contactless solar cell characterisation | Electrical and optical modelling of next generation solar cells | Signal processing
Suitable for recognition of Work Integrated Learning (industrial training)?
No
- Research environment
- Expected outcomes
- Supervisory team
- Reference material/links
- You'll join the ACDC group at SPREE, a collaborative PV lab advancing contactless characterisation and ML for PV applications.
- Day‑to‑day guidance from a postdoc (Zubair/Arthur), with senior oversight from Prof. Ziv Hameiri and access to the wider ACDC team.
- Access to our contactless Jsc equipment, high‑power computing hardware, and foundational Python repositories and codes.
- Onboarding to data and safety procedures, weekly check‑ins, and clear milestones aligned to ToRs project‑plan expectations.
- Develop a deeper understanding of contactless Jsc measurement approaches
- Design and develop a modelling workflow to synthesise next‑generation solar cell data for ML training
- Design and develop a clean, reproducible ML pipeline to accurately extract Jsc from lower‑resolution input data
