Wellington Faculty of Engineering continues successful bids for Marsden funding

Proposed research highlights the faculty’s commitment to high-quality research and work that can generate positive impact

Dr Qi Chen and Dr Fang-Lue Zhang from the School of Engineering and Computer Science and Dr Jianzhao Geng from Robinson Research Institute have received Fast-Start grants for research on genetic programming (GP), mixed reality, and controlling magnetic flux in superconducting magnets, respectively.

Titled Genetic Programming for Evolving Interpretable Models for Symbolic Regression, Dr Chen’s research will explore how artificial intelligence (AI) tools can learn and generate interpretable prediction models, such that people can understand how the predictions are made. “With increasing emphasis on achieving accountable and transparent AI, explainability and interpretability in modelling are emerging as crucial factors. This is especially true of high-risk environments, where a small error can have massive consequences – with disease prognosis, for instance” she says. “The beauty of GP is that it can positively influence building interpretable models across sectors, including aquaculture, climate change or underground water resources exploration, among others.“

Dr Zhang’s research, on Reconstructing Dynamic Panoramic Scenes in Mixed Reality, will involve developing a framework to rapidly process scenes captured from a moving camera, containing moving objects. “Today, 360-degree technology is being used extensively to capture and augment real-world 3D environments. But as current technologies are unable to handle dynamic scenes and objects well, interactions with the scenes are often lost,” he says. “Based on current industry trends, there will be an increasing demand for immersive experiences, and greater focus on virtual or mixed reality. And that’s one of the main reasons I chose this area – to better understand how we can leverage technology to enhance the capability of the current virtual reality system.”

Dr Geng’s earlier research in applied superconductivity, prior to joining Robinson Research Institute, focused on using inductive ways to generate DC current in superconducting magnets. His Fast-Start project, Ultra-precise control of magnetic flux quanta in high-Temperature superconducting magnets, will focus on ways to manipulate the motion of individual quantized fluxes, which are the minimum units required to form magnetic fields in high-temperature superconductors. “If our research can achieve the proposed method, we’ll be able to generate extremely accurate voltage and achieve stable currents and magnetic fields in superconducting magnets” says Geng. “We hope to use this method to develop stable power supply for MRI /NMR magnets.”

Dr Chen and Dr Geng are already involved in other projects that have received Marsden funding – Chen is part of Professor Mengjie Zhang’s Genetic Programming for Symbolic Regression project while Geng is working with Dr Chris Bumby and Professor Rod Badcock on investigating the irregular DC electric field generated by moving permanent magnets over a high-temperature superconducting film.

Each Fast-Start project will receive funding of $300,000 over a period of three years