Source | Getty Images
A novel-based AI manufacturing augmentation platform, supported by the Australian Composites Manufacturing CRC (ACM CRC, Sydney) and developed at Deakin University’s (Burwood, Australia) Applied AI Initiative, has shown the potential to create billions of dollars in efficiencies when adopted by composites manufacturers.
ForgeX.ai contains apps enabling manufacturers to use natural language queries to collect and search their data, run experiments on this to help optimize processes, and migrate and integrate data across enterprise resource planning (ERP) systems. It was supported by $110,143 in ACM CRC co-funding and will be made freely available to Australian manufacturers, beginning with those in the composites sector.
“The potential impact of allowing companies to ‘converse with their own factory’ cannot be overstated. And this version is just a chatbot interface. ForgeX.ai will be more powerful still if developed to offer agentic AI,” says Luke Preston, CEO at ACM CRC. “All the same, our estimates are that it could offer between 30-50% efficiency gains to manufacturers with no data system currently, and only a little less for those who do. If we could roll it out to just 10% of the addressable market in Australia, it would mean $4.1 billion in economic benefit. We also estimate that it can enable the introduction of Industry 4.0 at 1/100th the cost.”
The project lead, Deakin Applied AI Associate Professor Scott Barnett, says that the solution had origins in a previous (and discontinued) Australia’s Economic Accelerator-funded project using AI to reduce waste in manufacturing by identifying the root cause of defects. This was followed by exploring other use cases to help manufacturers who lacked software engineers to build in-house AI solutions.
“They all depend on software, but typically they don’t have the in-house capabilities,” explains Barnett. “So we went looking at ways to build tools that can help power up people who are perhaps traditionally not software engineers and use AI to bridge the gap as far as we can.”
ForgeX.ai is able to generate text, charts and images from users’ queries, and is built to be operated and calibrated by workers who have minimal training in machine learning. The system also focuses on protecting the IP of partners by keeping everything in Australia and on-premise if required. Data is always private and secure.
“It also allows all the content available in an organization to be easily and effectively structured,” adds Barnett. “We put something over the top of that to make everything more reliable and robust than Excel spreadsheets. So there will be less time spent on administration work, basically.”
“The platform is available now for partners and composite manufacturers, and we hope to expand beyond composite manufacturers next year,” says Preston. “If your business could benefit from this system, please reach out to ACM CRC.”