Ellen Phiddian
Cosmos science journalist
Researchers have used quantum physics and machine learning to quickly and accurately understand a mound of data – a technique, they say, could help extract meaning from gargantuan datasets.
Their method works on groundwater monitoring, and they’re trialling it on other fields like traffic management and medical imaging.
“Machine learning and artificial intelligence is a very powerful tool to look at datasets and extract features,” Dr Muhammad Usman, a quantum scientist at CSIRO, tells Cosmos.
“Quantum computing is an emerging technology, and it has the potential to offer immense computational capabilities.”
Quantum computing uses quantum properties, like superposition, to process information much faster than traditional “classical” computers.
“This special property allows you to process very, very large data sets very, very efficiently,” says Usman.
While usable quantum computers are still a few years away, they’re improving rapidly.
Usman and his team have simulated a quantum machine learning technique on classical computers to test it out. They trialled it on a vast dataset they had easy access to at the CSIRO, linked to groundwater monitoring.
“Groundwater monitoring is a very complex problem, and it is very costly,” he says.
Traditionally, it’s monitored by physically going to sites, collecting samples, and taking them to labs to look for signs of contamination. CSIRO has developed sensors which can monitor groundwater remotely.
“Those sensors provide large amounts of data to process and then determine the quality of the waters,” says Usman.
The team was able to process the flood of data coming from the sensors. They’ve described the technique in a paper published in Advanced Science.
But groundwater monitoring is just a start.
“The technology that we are developing, it has very broad implications,” says Usman.
The researchers are now trialling the same technique on medical imaging, and the Olympics. They’ve been awarded funds by the Queensland Government to develop technology for the 2032 Olympics.
“What the Queensland Government wants us to do is develop efficient quantum solutions that can be deployed in Brisbane Olympics, to make sure that traffic and management of large crowds can be efficiently done,” says Usman.
“Hopefully this will be the very first Olympics which will benefit from quantum.”
They’re also giving the technique a spin on MRI scans of injured knees.
“We applied this quantum technology to a large number of MRI scans of knees, and we showed that our technology works very well, and it can extract features from those images very efficiently, which might not be otherwise possible from conventional, existing technologies,” says Usman.
For now, the work is all proof-of-concept, but Usman hopes to see it used practically in the next few years.
“We are hoping to scale up the technology, to slowly move towards larger and larger and more complex datasets, and then gradually bring the technology into the market once we are ready,” he says.
“Our hope is that as the technology becomes mature in the next few years, these kind of solutions will start becoming more impactful in real-world applications.”
