Blog Archives

Dragonflies and Driveless Cars

Dragonfly

Despite the fact that dragonflies can’t drive cars, understanding how their brains work is improving selective attention for artificial vision systems, for applications such as driveless cars.

A recent study by Dr Steven Wiederman and published in eLife, demonstrated how dragonflies are highly efficient predators due to the highly complex nature of their brain. Specifically, cells in their brains, called Small Target Motion Detectors, can predict the direction and location of its prey.

Further understanding of such complex neurological systems can be applied to autonomous robots and driverless cars.

Articles:

Wiederman SD, Fabian JM, Dunbier JR & O-Carroll (2017) A Predictive Focus of Gain Modulation Encodes Target Trajectories in Insect Vision, eLife, 25th July, DOI: 10.7554/eLife.26478.002

Bagheri ZM, Cazzolato BS, Grainger S, O’Carroll DC & Wiederman SD (2017) An Autonomous Robot Inspired by Insect Neurphysiology Purses Moving Features in Natural Environments, Journal of Neural Engineering,13 July, DOI: 10.1088/1741-2552/aa776c

University of Adelaide media release

How the Dragonfly’s Surprisingly Complex Brain Makes it a Deadly Hunter, Gizmodo

Dragonfly Brains Have ‘Killer Cells’ That Can Predict the Movement of Their Victims – and It Could Lead To Robot Supervision, the Daily Mail

Dragonfly Brains Predict the Path of Their Prey, Science Daily

Top 100 Articles of 2016

Screen Shot 2017-03-30 at 9.47.27 pm

The publication “Fast machine-learning online optimisation of ultra-cold-atom experiments” was ranked in the top 100 articles published in Scientific Reports in 2016, receiving 11820 views.

The paper was the result of a collaboration between Australian National University and Professor Andre Luiten.

Scientific Reports is part of the Nature publishing group and more than 20000 articles were published in 2016.

prof-andre-luiten-ctd

Reference: Wigley et al (2016) “Fast Machine-Learning Online Optimization of Ultra-Cold-Atom Experiments” Scientific Reports, 6, 25890. doi:10.1038/srep25890