An agricultural robot prototype, developed by the University of Sydney’s Australian Centre for Field Robotics, has been demonstrated for the first time to the farming community at the recent Central West Farming Systems Precision Agriculture seminar at Condobolin, NSW.
Dubbed SwagBot, the machine has been specifically designed for Australia's livestock industry to keep an eye on livestock wellbeing, as well as analyse soil and pasture health where the stock are grazing.
Once commercialised, it will allow farmers to remotely monitor animal health, pasture biomass, take soil samples and check on weed growth from the home laptop or smart phone.
Weighing 190kg, SwagBot is electrically driven, has a high ground clearance, passive suspension and mid articulation, enabling it to detect and clamber over fallen logs.
University of Sydney senior engineer, Javier Martinez, says the aim of the project was to produce low-cost technologies to benefit Australian agriculture and horticulture.
"There are two platforms to come out of that – Digital FarmHand and the long-range, long-term automonous SwagBot," he says.
"We have a research project focused on using cameras to detect lameness in dairy cows with the outcomes feeding directly into the SwagBot project.
"The intention is SwagBot will go into the paddock on long-term missions of a few days. We aim to have solar powered recharge stations out in the field or eventually have solar panels on the robot."
Martinez says feedback would be gathered on how SwagBot would survive being rubbed or knocked by cattle.
"We have experience herding with the robot and there is a window where the cattle are terrified of it, and then it becomes part of the furniture," he says.
"We haven’t quite got to the point where they think it’s a scratching post."
"Their acceptance depends on the individual animals – how used they are to having strange things in their paddocks. We had one herd where they were comfortable with it by the end of the day."
Martinez says one of the early research projects, in collaboration with the University of Sydney’s departments of agriculture and veterinary sciences, was to determine how the cattle would react to the robot’s presence.
"When the cattle have a platform which is as agile and omni-directional as this, they don’t know what to expect," he says.
Under the next phase of development, SwagBot will use cameras and sensors to monitor pasture biomass and health.
"We aim to have a traffic light indication to say this pasture is ready for grazing, or for cattle to be moved to a fresh pasture," Martinez says.
"I want them to be comfortable enough with SwagBot so the robot takes the cattle to the best pasture or watering points, and they learn to follow it.
"We want to give farmers the extra information on pasture and soil current practices so they can look at where there might be improvements."
University of Sydney lead systems engineer, Matthew Truman, says SwagBot was a platform for integrating other technologies under trial.
"We have used cameras and computer vision in autonomous herding projects to track the cattle," he says.
"We are working on the user interface for SwagBot and it will mostly likely be a PC or mobile phone. It is built ruggedly, modular and low cost so parts can be replaced easily – the idea is to bring the cost down so it is economically viable for a farmer.
"The ultimate aim is to provide an economic benefit."
SwagBot has four independent driving wheels, making it highly manoeuvrable.
"It is not just a data collection platform but it can be doing physical tasks so it is equipped with a spray and drill unit," Truman says.
"It could use an on-board computer to detect and spray weeds, and collect soil samples for soil fertility mapping."
SwagBot can operate for six hours in the paddock before recharging, and is capable of travelling at 10km/h.
Field days are being used as a forum to collect farmer feedback on SwagBot.
Truman says the robot is still in the research and development phase, with commercialisation a few years away.
"This will be labour saving and give farmers the capability to collect daily data on pasture, livestock and soil fertility," Truman concludes.