Dr Nathan Kirchner

Innovation:
Toolbox Spotter

Chief Technical Officer, Laing O’Rourke/Presien; PhD (Robotics, Mechatronics, AI), University of Technology Sydney

The Toolbox Spotter is an innovation designed to improve the safety and productivity of heavy industries by using artificial intelligence (AI) vision systems.

There are roughly 200 workplace deaths and 100,000 serious workplace injuries per year in Australia, costing the nation an estimated $62 billion every year.

Being struck by an object is by far the leading cause of serious accidents, and heavy industries, such as construction, logistics, mining, manufacturing, and agriculture and forestry, disproportionately contribute to these statistics.

Analyses suggest that the vast majority of these accidents are due to distraction, lapses of attention, or failure to see a potential hazard.

Traditional approaches and technologies to safety in this sector have not been sufficient. That’s why Laing O’Rourke Chief Technical Officer Dr Nathan Kirchner built and led his team in developing a world-leading modular artificial intelligence computer vision system based on a deep convolutional neural network (D-CNN) that is optimised for heavy industries.

The AI sees objects, understands what they are and decides what actions to take in response to them seamlessly and in real-time.

The AI sees objects, understands what they are and decides what actions to take in response to them seamlessly and in real-time. In safety applications it is designed to eliminate the people-to-vehicle, vehicle-to-vehicle, and vehicle-to-infrastructure accidents that are all too common in heavy industries.

However, the technology is capable of much more. It is already being applied to a wider range of applications, including security, assurance, productivity, and reporting.

Among Kirchner’s innovations in devising the technology were building a world-leading high-tech team in a traditionally low-tech industry sector, securing funding for a long-term development project within a large corporation, taking fundamental research in D-CNN and translating it to a product, and training the AI on real-world situations.

With his team, Kirchner transformed the fundamental research into a form factor that was useful and adoptable by heavy industries. He is now the founder and CTO of Presien, the company spun-off to commercialise the technology.

Judges’ comments:

“An excellent initiative to support safety in these high-risk workplaces and a good practical application of artificial intelligence vision capabilities.”

MORE IN Mining, Oil & Gas

Kevin Chong

Turning data into intelligence

Professor Jun Huang

Reducing CO2 from oil refineries