Swagbot can herd cows, tow heavy trailers, and traverse rugged terrain and has been designed to manage livestock on Australia's vast sheep and cattle stations, which are often remote and difficult to access. A trial which began last month has confirmed that SwagBot is able to herd cattle, and can navigate its way around ditches, logs, swamps, and other features of a typical farm landscape. The next step will be to teach the robot how to identify animals that are sick or injured, says Salah Sukkarieh of the University of Sydney, who is leading the trial. "The trial has been very successful so far, so it's given us the confidence to move to the next phase," he says. His team is planning to fit the robot with temperature and motion sensors to detect changes in body temperature and walking gait, as well as colour and shape sensors to make sure the animals have enough pasture to graze on.

How often do you ring your bank and forget the special dates, places or names needed just to get through security? Apparently, it takes us 45 seconds on average just to confirm who we are. Now that might not seem very long, but if you're a global bank like Citi receiving 35 million calls a year in Asia alone, that adds up to 437,500 hours of staff time a year. But by using computers to identify our voices, this authentication process can be cut to 15 seconds on average, saving the bank pots of cash and us lots of hassle. Citi has just begun rolling out this kind of voice biometrics authentication for its 15 million Asian banking customers, starting in Taiwan, Australia, Hong Kong and Singapore.

The depths of the ocean can be a dangerous place, even for a drone. Incredible footage has captured the moment a Great white shark mistakes the machine for food and attacks it with amazing speed and precision. It is believed the video was filmed off the coast of Australia where Great whites are commonly spotted in southern waters from Exmouth in Western Australia to Southern Queensland. The video shows the metal drone sitting deep in the ocean, when almost out of nowhere the underwater beast appears. Within seconds the Great white wraps it teeth around drone and viciously tugs at it before realising the hunk of metal is not a tasty meal.

There's been an election in Australia, so once more, with all the regularity of a cuckoo clock, politicians and pundits alike are proposing that electronic voting is the answer. So, here we go again, explaining why it's a bad idea. First, if e-voting is the answer, what is the actual question? "We're a grown-up democracy, it shouldn't take eight days to find out who's won," said Malcolm Turnbull, who is likely to continue as prime minister. "[E-voting is] something we must look at", he said.

Discussing the pros and cons of articial intelligence is, to be sure, an emotional topic. People have feared artificial intelligence (AI) almost as soon as it was invented. Hollywood in particular has done a masterful job of stoking those fears with movies like "2001: A Space Odyssey," "Terminator" and "The Matrix" all making AI systems into demonic forces. Most recently, renowned scientist Stephen Hawking told Larry King he thought humans and AI could "co-exist," but that "a rogue AI could be difficult to stop" without appropriate safeguards in place. Coming from someone like Hawking, that kind of warning carries some weight with people.

Nature can make our most advanced technology seem clunky and basic. Picture a drone attempting to navigate a pile of rubble during a search and rescue mission: it would likely bump in to a wall or two along the way, even if it were supervised by a human pilot. In the same setting, an insect would gracefully swoop past obstacles in order to reach its destination. A research group at the University of Queensland, Australia, is studying the flying techniques of birds and bees in order to develop new navigation systems for Unmanned Aerial Vehicles (UAVs). Engineers have spent years perfecting advanced sensors, radar, and GPS systems that help drones identify and reach their targets.

An artificial intelligence platform, called Rechtwijzer, could soon give lawyers in Australia a run for their money by being called on in legal battlegrounds like divorce, custody, employment and debt disputes. National Legal Aid and RMIT University are showcasing Dutch technology that could make time-consuming and expensive court conflict a thing of the past. The technology would be similar to eBay's dispute resolution service that helps people log on, rather than lawyer up. The dispute resolution robot was born in the Netherlands and can mediate everything from divorces, tenancy disputes, and employment, debt and consumer matters. For custody matters, for example, it will ask the ages of the children to be sensitive to their development needs.

The skies above are growing increasingly crowded with satellites zipping across Earth's upper atmosphere, relaying signals for everything from the picture on your television to the map on your phone. While this constant communication makes the world go round, all of that orbiting technology brings with it a problem, in the form of space junk – the debris from rocket launches and defunct satellites which hangs on in space. Lapping the Earth at thousands of miles per hour means even smallest chunks of metal or flecks of paint can cause significant damage if they run into the path of a satellite. But scientists at the University of Surrey are gearing up to test technologies which could target potentially hazardous space junk and remove it from orbit before it can cause any damage. The first experiment will use a net to capture a target in the form of a small CubeSat launched from the main satellite.

A system of nested dichotomies is a method of decomposing a multi-class problem into a collection of binary problems. Such a system recursively splits the set of classes into two subsets, and trains a binary classifier to distinguish between each subset. Even though ensembles of nested dichotomies with random structure have been shown to perform well in practice, using a more sophisticated class subset selection method can be used to improve classification accuracy. We investigate an approach to this problem called random-pair selection, and evaluate its effectiveness compared to other published methods of subset selection. We show that our method outperforms other methods in many cases when forming ensembles of nested dichotomies, and is at least on par in all other cases.

Do you believe that companies live and die by their understanding of how customers use and grow with their products? Do you love the idea of rapidly iterating on new features as experiments against thousands of users to find the ones which move the needle and hit customer engagement out of the ballpark? Would you like to drive real change at the world's hottest SaaS company, and the best place to work in Australia? Atlassian is looking for an analytical mind to work in our Product Growth team to analyze the way customers use our our amazing suite of products – JIRA, Confluence, BitBucket, Bamboo, HipChat... We use real data science to point the way, then design experiments to improve customer engagement and growth; we work with Product Managers, Designers and Engineers to build a magical experience for customers. It's not for the faint of heart -- you'll need to slice and dice large-scale data, looking for key trends and patterns.