CSIRO researchers have demonstrated the practical application of quantum machine learning, showing how quantum computing can enhance the analysis of large datasets across multiple industries, including traffic management, agriculture, healthcare, and energy optimisation.
By harnessing the properties of quantum computing—superposition and entanglement—the team compressed and analysed vast amounts of data with efficiency and accuracy beyond the capabilities of traditional computing systems.
Unlike conventional binary computing, which processes data in a linear fashion, quantum bits (qubits) can exist in multiple states simultaneously, enabling parallel processing of complex problems, researchers said in a news release.
Dr Muhammad Usman, a CSIRO quantum scientist and senior author of the study, said the research proved quantum machine learning can simplify large datasets while retaining critical information.
“With the global volume of data doubling every few years, quantum computing’s ability to handle this complexity will become increasingly valuable,” Dr Usman said.
“Our work focused on groundwater monitoring as a case study, but quantum machine learning has broad applications in any field requiring fast, detailed analysis of large datasets.
“As practical applications for machine learning rapidly increase, we expect that integrating the tremendous computational power of quantum in machine learning will offer transformative impact in solving many industrial and real-world problems.
“For example, this could transform how we optimise traffic routes to minimise congestion on roads and reduce harmful emissions or process medical imaging with unprecedented accuracy to enable fast and reliable diagnosis.”
The findings come as UNESCO has declared 2025 the International Year of Quantum Science and Technology, highlighting the growing global emphasis on advancing quantum research and its applications.
Dr Liming Zhu, Research Director at CSIRO’s Data61, said the breakthrough strengthens confidence in quantum machine learning and helps define key performance metrics for future advancements.
“CSIRO’s breakthrough not only builds confidence in the benefits of quantum machine learning but also serves as a guidepost. By identifying key application performance metrics and challenges, our work helps shape the trajectory of hardware and software innovation, bringing us closer to real-world demonstrations using quantum,” Dr Zhu said.
“UNESCO’s International Year of Quantum Science and Technology provides us with a great opportunity to promote the valuable work our scientists do as well as help others to better understand this complex field.
“Australia has been a world leader in quantum technology research and development for almost 30 years and this work adds to the pool of significant local innovations.”
The study, titled ‘Self-Adaptive Quantum Kernel Principal Component Analysis for Compact Readout of Chemiresistive Sensor Arrays,’ was published in the high-impact journal Advanced Science.
It was co-authored by CSIRO researchers Dr Zeheng Wang, Dr Timothy van der Laan, and Dr Muhammad Usman.
