Dr Nigel Greenwood, ilab’s Germinate 6 accelerator graduate, was last night awarded ..
out of 100 finalists in the Anthill Top 100 Cool Company list
WINNER Anthill Cool Company – Innovation Category
Nigel is currently domiciled at ilab with his 32 CPU “Origin” power PC, building his Turbine Machine Genes international startup business, using complex systems theory to understand really complex jet engine performance thereby rendering neural network theory obsolete.
This follows last win by Monica Davis, founder of Rumbl (ilab Germinate 6) as the Young Innovator category of the nation The Australian Innovation Challenge award, and the week before Brad Parsons, of MOVUS (ilab Germinate 6) taken out the CSIRO Collaboration prize at the recent Tech23 pitch comp.
Following, in his own cool words, is what Nigel’s Turbine Machine Genes and EMI IP holding company get up to..
EMI is a pure R&D entity. Its commercial spin-off, Turbine MachineGenes Pty Ltd (‘TMG’) commercialises the technologies EMI develops in the context of complex, high-value engines, worldwide. TMG is a graduate of ilab Germinate accelerator.
Using a novel proprietary platform of evolutionary machine-intelligent algorithms, EMI and TMG are able to evolve—quite literally, evolve— sophisticated mathematical models of complex dynamical systems from noise-polluted partial information, and then use these models for machine-intelligent analysis and control. Putting it simply, they enable machines to understand complex systems (including other machines).
The Founder of both companies, Dr Nigel Greenwood, has developed this proprietary technology, a new form of machine-intelligent algorithm called “phi-Textured Evolutionary Algorithms” (phi-TEA), with a patent application at PCT stage. As of September 2015 the International Examiner has ruled all 70 claims are globally “novel and inventive”, so a large number of basic patents in artificial intelligence are anticipated internationally.
His motivation was that he’d been working in machine intelligence for two decades since the end of his PhD in applied mathematics, developing a new platform technology enabling complex systems to be reconstructed from partial data. After successfully simulation-testing this technology in medical and military applications, he realised there was a global need for machines to be able to “understand” and control complex systems (including other machines) in a way easily communicable to human operators and suitable for mission-critical applications, such as medical devices and turbine engine maintenance.
Previously it has been extremely difficult or impossible to understand exactly what is happening within a complex dynamical system like a turbine engine in operation. Small changes in engine thermodynamics and physical dynamics can rapidly cause it to become idiosyncratic. EMI/TMG answers the questions: without dismantling the engine or stopping it, what is happening; does it need intervention; and if so, what?
Phi-TEA is expected to reduce significantly the risk and cost associated with turbine engine maintenance, and in so doing transform the engine maintenance industry worldwide. Aviation engines on average cost about $US 16.25 million each. Engine failure and maintenance are very expensive: an engine in the hangar means a grounded aeroplane; a failed/damaged engine means the manufacturer must replace it.
This has impressed potential customers throughout the world. Their first client, one of the world’s top aviation engine manufacturers, decided to fund a demonstration project after reading the patent application and seeing Dr Greenwood’s demonstration of his platform technology in a medical context.
Dr Greenwood plans to make neural networks– the current global gold standard for machine-intelligent technologies– obsolete for a large class of applications.
In summary, EMI/TMG intends to overthrow the current international gold-standard technology with a new, proprietary technology, protected by patents-pending. We are about to transform the capabilities of global high-technology industries and generate new IP. And we’re going to do all this via the Cloud, so can do it from a computer laboratory in suburban Brisbane.