Dragonfly technology provides pinpoint accuracy and super speed

Dragonfly technology provides pinpoint accuracy and super speed

Dragonflies are one of nature’s most effective predators. They can target, chase and capture their prey with an amazing 95 per cent hit rate.

Our researchers were first to discover how the dragonfly can focus on a single moving target, shutting out all distractions.

Now we have found a way to adapt dragonfly nerve cell functions to machine learning technology, with multiple applications in defence industries.

Our computer scientists, neuroscientists and mechanical engineers combined forces to develop a unique algorithm that copies the insect’s visual tracking ability.

Virtual reality testing has now shown that this autonomous pursuit algorithm runs 20 times faster than comparable algorithms developed elsewhere, while matching their accuracy. This means it requires far less relative processing power and is far more efficient.

The alogorithm has already been put to good use by mechanical engineering researchers developing autonomous pursuit robots.

Lead researcher Dr Steven Wiederman’s team has also emulated the dragonfly’s ability to predict where its prey will travel, which enables it to set up an ambush.

This has led to further collaboration with the University’s of Adelaide’s Australian Institute for Machine Learning to develop drone-tracking systems.

“We’re excited to further define the principles underlying neuronal processing,” Dr Wiederman says.

“Translating them into advanced artificial vision systems could result in some incredibly effective autonomous robotics and drones.”

He believes there are many more potential applications for our innovative technology, including neuronal prosthetics to improve the lives of people with brain impairments or damaged nervous systems.

“The possibilities are revolutionary,” according to robotic vision specialist Professor Ian Reid.

“Artificial neural networks, together with vast computing power and data volume, have enabled step-change in the level of intelligence machine learning can achieve.”

Bringing together researchers from diverse fields has multiplied the benefits by compounding our expertise and opening up new opportunities.

Tagged in Defence, cyber and space, defence, machine learning, autonomous, drones, robotics