Towards a Smart PCR Process: boosting degraded DNA with machine learning in real-time
Who Professor Adrian Linacre from Towards a Smart PCR Process
What Towards a Smart PCR Process developed a DNA amplification system that uses real-time feedback and machine learning to adjust the process as it runs. This improves the quality of genetic data from degraded or low-level samples, increasing the chances of recovering usable results for forensic investigations and other scientific applications.
Winner of the 2025 University of Technology Sydney and Australian Federal Police Eureka Prize for Excellence in Forensic Science.
Towards a Smart PCR Process accepting the 2025 University of Technology Sydney and Australian Federal Police Eureka Prize for Excellence in Forensic Science. Photo by Getty
Image: Wendell Teodoro/Getty Images for Australian Museum© 2025 Getty Images
Towards a Smart PCR Process has developed a DNA amplification system that increases the chances of recovering usable results from low-level samples. What is the impact of this for forensic investigations?
Very many of the samples collected from crime scenes have trace levels of DNA – insufficient to generate useful genetic data. Many also have substances that inhibit the generation of DNA data. This generation of DNA data relies on the PCR process. Our smart PCR concept is designed to transition those samples that currently do not generate DNA to ones that do. The implication for the criminal justice system is exoneration of the innocent and inclusion of fewer persons by chance.
Aside from forensic investigations, what other scientific applications might this your system be useful for?
These are actually much bigger areas where our technology can help. Any process that uses the PCR method will benefit from this concept. So much of medical sciences are now reliant of the PCR process. Screening for any genetic change, not only in medical science but in agriculture and pharmaceutical sciences, will benefit from this technology: so, the scale of new applications is immeasurable.
What challenges did you face in your discipline that led you to develop this technology?
The biggest challenge is that there is currently no machine capable of performing this work.
The second big challenge is that methods in forensic science go through extensive validation steps and must be robust, reliable and repeatable: as a consequence of our proposed method, the outcome will not be repeatable.
This DNA amplification system uses real-time feedback and machine learning. How does this work, and is this approach becoming more common in your discipline?
This is inevitable. The application of AI and machine learning offers immense benefits and has already been used in reading outputs from machines to remove the human element in reading printouts. We envisage another step, although a big step, in the introduction of AI to forensic science.
“Screening for any genetic change, not only in medical science but in agriculture and pharmaceutical sciences, will benefit from this technology: so the scale of new applications is immeasurable.”
What is next for your team?
There are IP issues very much part of the next step – but we as the inventors really just want to see this happen!
Towards a Smart PCR Process at the 2025 Australian Museum Eureka Prizes Award Ceremony. Photo by Mel Koutchavlis
Image: Mel Koutchavlis© Mel Koutchavlis
What does winning the inaugural Eureka Prize for Excellence in Forensic Science mean to you?
It is a great honour. Firstly, it is excellent to see a Eureka Prize in the forensic sciences. Secondly, it is marvellous that an application that can change greatly the forensic process was chosen as a winner, and we hope that this sets the standard for future winners.
The Australian Museum Eureka Prizes are the country’s most comprehensive national science awards, honouring excellence across the areas of research & innovation, leadership, science engagement, and school science.
