Sparc Technologies (ASX: SPN) has partnered with the University of Adelaide’s Australian Institute for Machine Learning (AIML) to jointly develop AI software aimed at modernizing the performance assessment of protective coatings.
This project focuses on applying computer vision and machine learning to improve the efficiency, accuracy, and consistency of corrosion assessment methods that have remained largely unchanged for over 25 years.
The pilot program has already demonstrated a proof of concept based on ISO 12944 corrosion boundary testing, providing early technical validation of the approach.
The software is developed using large historical datasets to train models that can achieve higher accuracy and consistency than manual evaluation.
In addition to reducing turnaround time, AI approaches enable the capture and reporting of large numbers of data points, allowing for more comprehensive statistical analysis than a single manual measurement.
Performance evaluation of coating modernization
Testing of protective coatings typically relies on scribe-based methods that intentionally damage the coating to promote corrosion and evaluate performance by measuring corrosion creep from the scribe.
Current evaluation techniques rely on subjective human judgment and manual measurements, making the process laborious and time-consuming, and subject to operator-to-operator variability.
Sparc and AIML’s software is designed to replace this process with visually trained AI models that can automatically detect corrosion boundaries and coating delamination.
The company estimates that its AI-based workflow can reduce evaluation time from about 40 minutes to tens of seconds per result.
Commercial opportunities and industry support
Sparc estimates that approximately 850 laboratories are available across coatings industry, testing, and research organizations worldwide to perform relevant end-user performance evaluations.
The company has received letters of support from multiple coatings industry stakeholders, reflecting early interest in commercial applications of this technology.
The company plans to pursue a commercialization route based on industry collaboration and future software licenses to established testing laboratories and coating companies.
Beta testing in third-party labs is targeted for no more than 12 months.
Adjusting the focus of innovation
Managing director Nick O’Loughlin said the partnership is consistent with Sparc’s focus on innovation in the protective coatings space.
“This partnership with the University of Adelaide’s Australian Machine Learning Institute leverages the complementary skills of both teams to develop AI-based software for the protective coatings industry,” Mr O’Loughlin said.
“It has the potential to significantly improve and accelerate research and development under test methods that have not changed in more than 25 years.”
Professor Simon Lucey, Director of AIML, added: “Replacing outdated manual corrosion assessments with computer vision has the potential to increase efficiency and improve outcomes for the coatings industry.”
