By David Sly
Artificial intelligence and machine learning are the great engines of change driving the Fourth Industrial Revolution – or Industry 4.0, as it is commonly called – and Adelaide has quickly positioned itself as an important hub for helping businesses implement the great challenge to adapt and innovate.
This has resulted in local companies automating tasks from screening for urinary tract infections to monitoring pests inside horticultural greenhouses, using innovative computer programs and machines that have been invented locally.
The Australian Institute for Machine Learning (AIML), located on North Terrace but soon to be relocated within the Lot Fourteen innovation hub (the former Royal Adelaide Hospital site), offers South Australian businesses the chance to be early machine learning adaptors.
Having emerged in early 2018 from the University of Adelaide’s Australian Centre for Visual Technologies (ACTV), led by global expert in video semantics Professor Anton van den Hengel, the AIML has received $7.1 million funding from the State Government (in addition to $5 million invested by the University of Adelaide) to work on such programs as improving traffic flow, defence projects and assisting the progression of emerging small and medium-sized enterprises.
The institute’s impact and influence is growing rapidly. In the decade since ACTV began and morphed into AIML, the team has grown from five people to more than 100, including the brightest minds emerging from the university’s mathematics and computer science courses.
While a third of AIML’s work pursues pure research, which has it currently ranked third in the world in computer vision, it also focuses on applying research outcomes to industry needs.
“This research is keeping us ahead of the curve. Technology has a six-month cycle before it takes the next leap – it truly is moving that quickly,” explains AIML business development manager Paul Dalby. “Most organisations can’t afford to staff this, so we have the pool of talent they can draw upon.”
AIML researchers and technicians are trying to teach machines new ways of solving complex manufacturing and industrial problems, training computers how to do tasks on their own through deep learning programs, and implement faster ways of completing tasks.
LBT Innovations was among the first SA-based businesses to achieve commercial outcomes from working with AIML. The company, which focuses on clinical microbiology, initially went to ACVT in 2010 with the idea of automating the reading of inoculated petri dishes while screening for urinary tract infections.
This is a time-consuming task for scientific specialists that mostly registered negative readings, and after extensive research, testing and trials, LBT had its own patented machine (the Automated Plate Assessment System) ready for international sale by the end of 2018.
“It’s not a speedy process to develop an original idea – creating and testing something that is not an off-the-shelf algorithm – but now we have arrived first in this part of the global healthcare market,” says Brent Barnes, CEO of LBT Innovations. “It’s great that we have been able to initiate this world first from Adelaide.”
Machine Learning can also take effect at more simplistic levels for businesses. AIML has supported graduate Jordan Yeomans to launch his company Advanced Innovations, which is helping farmers growing vegetables in greenhouses at Virginia, just north of Adelaide.
He has switched their pest monitoring from a manual to automated system, using smartphone links to custom-designed AI computer software.
The success of this process – which Jordan estimates saves about 20 hours of manual labor a week – will provide the springboard for Advanced Innovations to work with greenhouses around Australia.
Other early adopters taking advantage of this homegrown expertise include Maptek, Sydac and Signostics. AIML will also reinforce and improve leading SA organisations from SAHMRI to the renewable energy sector, ensuring they remain at the forefront of global attention and success.
However, beyond these success stories, more needs to be done and with urgency – 80% of small and medium-sized businesses in Australia are delaying adaptation of machine learning, while overseas comparisons show that the rush to make sustained investments in AI and machine learning has commenced in earnest.
“What AIML offers is a huge leg-up for SA business,” says Paul Dalby, “but at the moment, not enough are taking up this opportunity. We are doing more work with companies in Sydney and Melbourne, so we need more mid-sized companies to get on board and build now.”
AIML is offering training programs for company CEOs to be introduced to machine learning concepts and possibilities.
The first free event for this year is being held on April 10 at the University of Adelaide’s Nexus Building on Pulteney Street. (Register here).
“We want to initiate the conversation with SA business leaders to discuss what is possible, and what steps can be taken for their enterprises to move beyond existing boundaries,” says Paul.
“This is very important, because if some companies don’t take swift action in this very dynamic era of change, their business models could very quickly become obsolete.”
Visit I Choose SA to meet the people building business and industry in SA, and to find out how your choices make a difference to our state.