Unmanned Systems Technology 006 | ECA Inspector Mk2 USV | Antenna systems | Northwest UAV NW-44 | Unmanned ground vehicles | Navigation systems | Lunar X challenge

57 BHP Billiton, which is moving to fully autonomous operation by deploying test trucks for coal mining in New Mexico and a fleet of six for ore mining at Jimblebar, again in Western Australia, and the nearby Wheelara deposit, and is seeing a 12% increase in productivity as a result. Automated Systems Inc (ASI) in the US is highlighting the need for additional sensors to help automated mining and construction vehicles. It found that the inbuilt safety software in the trucks was detecting problems and then automatically shutting the trucks down without informing the remote operator why they had done so, forcing the operator to come out from the control centre to restart the vehicle and sort out the problem. To resolve this type of issue, ASI installs camera feeds so that remote operators can visually survey the area surrounding the unmanned vehicle to diagnose the stoppage. Additionally, ASI engineers are working on sensor technology that roots out false positives in obstacle detection algorithms as well as dynamic obstacle avoidance software that can safely navigate around unexpected hazards without the intervention of a remote operator. Agriculture Autonomous systems are also increasingly showing promise for agriculture by combining image sensors and machine learning algorithms. For example, the Bonirob agricultural robot has been developed by Bosch start- up Deepfield Robotics in Germany to identify and remove weeds within a field of crops without using pesticides or weedkiller. “We are using our expertise in sensor technology, algorithms, and image recognition to make a contribution to improving quality of life,” according to Deepfield’s general manager Professor Amos Albert. The robot is the size of a small car and uses both video and Lidar to determine its position alongside GPS navigation to find its way around fields with an accuracy to the nearest centimetre. This is used to automatically analyse different strains of plant being grown in a field to speed up trials of new strains. “Algorithms analyse the photos taken by scanners and cameras. This automatic screening saves a lot of time and effort,” says Albert. Analysis of the shape of a leaf from an image, coupled with machine learning, can also lead to identifying weeds that are then eliminated by a piston from Bonirob stamping them into the ground. “The leaves of carrots and chamomile, for example, are very similar in their early stages,” says Albert. “Over time though, based on parameters such as leaf colour, shape, and size, Bonirob learns how to differentiate more and more accurately between the plants we want and those we don’t.” Unmanned ground vehicles | Insight Unmanned Systems Technology | February/March 2016 Manufacturer Customer Location Fleet Komatsu Rio Tinto Australia 69 Caterpillar BHP Billiton Australia 12 Caterpillar BHP Billiton US 6 Caterpillar Fortescue Metals Australia 12 ASI Anglo America Australia, US Trials Unmanned transportation in mining This agricultural robot has been developed to identify and remove weeds without using pesticides or weedkiller (Courtesy of Deepfield Robotics/Bosch) Over time, using parameters such as leaf colour and shape, Bonirob learns how to differentiate between plants we want and those we don’t