Australasian Science: Australia's authority on science since 1938

Farms to Benefit from Robotics and Pest Surveillance

Robotics expert Prof Salah Sukkarieh of The University of Sydney has been named the Researcher of the Year by the Australian vegetable industry’s peak body Ausveg for his work on intelligent farm robots. His “Ladybird” was designed and built specifically for autonomous farm surveillance, mapping, classification and detection for a variety of different vegetables.

Sukkarieh says the automation of on-farm processes will help to increase efficiency and yield by having many of the manual tasks of farming performed by specially designed agricultural robotic devices. “Ladybird focuses on broadacre agriculture and is solar electric-powered. It has an array of sensors for detecting vegetable growth and pest species, either plant or animal. She also has a robotic arm for the purposes of removing weeds as well as the potential for autonomous harvesting.”

Sukkarieh says the Ladybird ‘s first field trip in Cowra was a success. The solar electric-powered bot was charged before heading to the onion, beetroot and spinach farms of Cowra, and was fully operational for 3 consecutive days on the farm. “The robot was able to drive fully autonomously up and down rows and from one row to the next while gathering sensor data. Sensors include lasers, cameras and hyperspectral cameras.”

Future testing of the Ladybird will include a robot manipulator arm under the vehicle that has potential for spot sensing or spot sampling. Automated harvesting is also on the agenda.

In separate research, the Invasive Animals Cooperative Research Centre (IACRC) is developing surveillance systems to enable farmers to detect and manage incursions by wild dogs. “Our team has developed proof-of-concept software that can automatically screen camera trap images from the field and recognise different dogs from facial features,” said Andreas Glanznig of the IACRC.

The IACRC team has been working on this technology across several pests to differentiate between species at trap and baiting stations. “Baits could be deployed from a device that recognises the target pest,” Glanznig said. “We could also develop technology that automatically notifies you of a trap capture as well as what’s in the trap,” he said.