Use your social profile entry

Robots and the future of agriculture: Australian experience

An element of automation already exists in many areas of agriculture. Precision agriculture, autosteer and controlled traffic farming are technologies enabled by global positioning systems (GPS). GPS is already the norm on many farms. Applied to traditional farming equipment such as tractors, harvesters, ploughs and sprayers, it has enhanced labour efficiency and helped curb costly waste by enabling large-scale crop farmers to harvest and spray fields with pesticide and herbicide with centimetre accuracy.

Robots are the next step, using these and other new technologies in combination with remotely-sensed data, to support new levels of decision making and assume some traditional farmer roles. Robots that can plant, fertilise, spray, weed, monitor and, ultimately, harvest, pack and transport crops will inhabit the countryside. While the drones hum overhead other remotely controlled intelligent surface, robots will be able to inspect and herd farm animals[1].

Robotic platforms[2]

Earlier this year at a commercial vegetable farm in Cowra, the ACFR (Australian Centre for Field Robotics) unveiled one of its latest agbot creations, the Ladybird, a sophisticated autonomous robot powered by solar energy that is capable of undertaking a variety of monitoring and evaluation tasks. It’s also an efficient killer: as it slowly wheels its way around a paddock it can locate and eliminate individual weeds by means of a targeted sprayer. Machine is equipped with sensors for measuring vegetable growth, and for the detection of animal and plant pests. It has a robotic arm that can remove weeds and potentially harvest vegetables.

But Ladybird is just one of many robotic platforms under development at the Centre, including a variety of drones and a pair of autonomous rovers christened Mantis and Shrimp.


These vehicles are excellent, particularly for tree crop applications. The robots are equipped with a whole number of different sensors that really span across the electromagnetic spectrum, using lasers to build three-dimensional pictures of the world. ACFR uses these in tree crop applications where the robots can drive themselves up and down the rows of an orchard, and using all of these sensors it can put together a really complete picture of the trees and how they are growing.

ACFR can automatically detect and count flowers and fruit, and give maps of that information to growers. That information turns out to be really useful for a decision support capability. Growers can look at the information covering their orchard and they can identify trouble spots and work out how to address those problems so that they can get really the maximum output from their acreage.

There are also major environmental benefits to be had from agbot use through a reduction in herbicide usage.

Because agbots like Mantis, Shrimp and Ladybird adopt a targeted approach to weed destruction, the overall use of agricultural poisons can be significantly reduced—by up to 40 per cent.

And with weeds in Australia developing a chemical resistance, a new emphasis on the use of microwave technology for weed control could potentially replace herbicides altogether.

It’s not unreasonable to imagine that in the future with technology like this you might actually be able to have mainstream organic produce that is actually farmed at the kind of scale that the current non-organic crops are farmed at.


The use of small-scale robotics also opens up the possibility of increasing the overall amount of productive land.

SwarmFarm has been developing a small robotic weed-killer called Swarmbot. The robot isn’t much to look at – it’s a fairly simple metal platform on wheels – but like Ladybird, Shrimp and Mantis, its simplicity belies its sophistication. It’s about lightweight machines that can travel very slowly down to a crawling speed, even stop at individual plants and go and intervene and manipulate those plants. In some cases the actual speed of the vehicle, of the robot, may be a lot slower than the speed of the full-sized tractors, but then you make up for that in terms of the fact that it can run overnight and also the fact that you can scale up or scale down these systems based on how many of the individual smaller platforms you have.

Aerial robots[3]

Another approach to counter the drawbacks of large tractors is to employ small aerial robots equipped with sensors.

Although large manned aircraft may be cost prohibitive for routine information gathering, small autonomous platforms have strong potential. ACFR has completed several projects where they developed and demonstrated aerial robotic systems for weed maintenance in an environmental monitoring context, including aquatic weeds such as alligator weed and larger woody weeds such as prickly acacia.

Examples of the some of the robotic platforms used are shown in Photos below. In these projects, the idea is to locate sparse concentrations of weeds that exist in large areas, and then to deploy herbicide locally and in a targeted manner. Weeds are automatically identified using classification algorithms that operate on visual imagery collected by the aerial robots. Herbicide can then be delivered manually or via a specially equipped robot. In the broadacre context, this type of approach can complement ground robot systems by rapidly finding concentrations of problem weeds that can then be efficiently targeted by the ground robots on an as-needed basis.

Robotic aircraft used in large scale farm enterprises for detecting weeds (top), and an example of a small robotic aircraft for doing the same thing but in small scale applications (bottom).

Robotic milking systems[5]

Automated or robotic milking systems address the twin challenges faced by dairy farmers of labour intensity and labour shortages. Automation reduces the need for people to be present during milking, providing additional time and flexibility to focus on other areas of farm management.

Robotic milking systems use sensors and a camera or laser-guided milking cups to guide a robotic arm to the udder, clean the teats and attach milking cups. When milking is complete the cups are removed automatically. Additionally, each cow may be fitted with a transponder, which registers the animal as it enters the dairy and collects data enabling milk production, milk quality and herd health to be monitored.

In Europe, manufacturers of automatic milking systems predict that by 2025 almost half of the herds in leading dairy countries, such as Denmark and the Netherlands, will be milked by robots. Robotic dairies, with varying levels of automation, have been available commercially in Australia since 2001. However, Dairy Australia believes adoption has been slow due to a lack of long-term performance information about robotic dairies in Australian pasture-based systems. While the cost of installing a robotic dairy is similar to installing a conventional dairy, the outlay remains significant and likely not warranted if an existing dairy is still functional and suited to the scale of the farming operation.

Agbot II[4]

The Queensland University of Technology (QUT) has designed and built a robot that could save Australia’s farm sector $1.3 billion a year by slashing the costs of weeding crops by as much as 90%.

This fully-autonomous agricultural robot, developed with support from the Queensland Government, is called Agbot II.

Agbot is equipped with sensors, software and other electronics that allow it to detect and classify weeds and then kill them either mechanically or chemically. In future versions, the robots could also feed back data on such things as soil and crop health and the state of diseases as they conduct their operations. This would enable better management decisions driven by paddock specific real-time information.

To date, Agbot II is concentrated on the three weeds that are relevant to Queensland: volunteer cotton, sow thistle and wild oats, and the vision system operated with 99% accuracy in the classification of the correct species based on the images collected by the robot cameras.

The light weight of AgBot II, which is about 600kg, will help reduce soil compaction that affects the yield by limiting the root development of the crops. Also due to weight, the robots can be deployed faster onto fields after rain to keep a tight control of weeds before they drop their seeds.

AgBots are also designed to work in groups, which increases the reliability of weeding operations. If one robot has a problem and fails, the others continue operating.

Another important characteristic of the Agbot ll is that it is solar powered, which makes it eco-friendly and easier on the farmer’s budget.

Challenges for adoption[5]

The main challenge to the adoption of robots in Australian agriculture is the fact that the technology is not sufficiently advanced for widespread uptake. In 2016, most agricultural robots for operation on individual farms are still in the trial or proof of concept stage.

The Queensland University of Technology identified four general barriers to widespread adoption of robots in agriculture in 2014.

  • Technology — Most agricultural robots are working in trial situations where they have continual access to technicians and engineering support. The challenge is to build a robot for commercial operations that is reliable and robust, simple enough to operate and low cost. A lack of coordinated data and standards may also prevent on-farm adoption, and there is a need for robots to be able to operate across multiple hardware and software platforms.
  • Regulation and certification — Protocols for the operation of autonomous robots may need to be developed, particularly in relation to safety on farms. Preliminary regulations for the operation of UAVs have been developed, however ground-based robots have not been included. Including on-farm robotic functions in such regulations may also affect insurance cover and premiums.
  • Business — As at 2016, there is not enough commercial performance data to identify the value or expected return on investment of robotics to Australian farm businesses. Further, agriculture consists of many small businesses with limited access to capital for business development, which may delay the uptake of robotics in the industry. This is in contrast to large mining or stevedoring businesses with financial capacity to invest in research and capital upgrades. The challenge for agricultural engineers is to develop low-cost technology without sacrificing quality or reliability.
  • Legal and socio-economical aspects — Liability for the actions of robots may need to be determined, particularly if robots acting autonomously cause damage to people, property, crops or the environment. The social acceptability of robots in the landscape may also need to be considered.

Challenges for the adoption of robots, partially addressed in the list above but likely to be significant, include the robustness of the technology for agricultural applications and the aging target user group.

Source –

Based on materials:





[5] Rural Industries Research & Development Corporation publication no. 16/033, Robots, 2016

Search tags: