Precision Ag Down Under

Precision Ag Down Under

Queensland farmer Andrew Bates is helping drive the development of robots suited to grain production activities via his business SwarmFarm.

Queensland farmer Andrew Bates is helping drive the development of robots suited to grain production activities via his business SwarmFarm.


When it comes to new technology, Australia is acknowledged as a country of early adopters and farmers are no different. Indeed, Australian agriculture was at the forefront of the precision farming revolution with the development of the Beeline guidance system back in 1998.

Today, about 80% of Australia’s grain growers have some form of autosteer and guidance with many running +/-2 cm RTK systems.

Autosteer and guidance is also becoming more widely adopted across other field scale industries such as sugar, cotton and rice as well as in orchards, vineyards and horticulture.

With guidance rapidly becoming a standard rather than an optional extra we are seeing farmers and researchers looking to expand the value of spatial management tools.

Yield, Quality Mapping

Many Australian production systems are driven by yield and to a lesser extent quality, so spatial yield data provides a vital data layer for spatial crop management.

Until recently the collection of spatial yield data has been limited to grain. Now yield monitors for potatoes, carrots, almonds, grapes and most recently field tomatoes have all been developed by Bernd Kleinlagel and his South Australian based company Advanced Technology Viticulture (ATV). Indeed, this season his yield monitors were fitted to all harvesters on California’s largest grape grower.

Export hay is an important rotation crop in many Australian grain farms but this crop left a hole in the collection of yield data. Western Australian-based Nick Ross, Precision Agronomics Australia, and local farmer Paul Hicks have now developed a hay yield monitor based on load sensors located on the mower conditioner. They found this approach to be more reliable than measuring bale weight.

While many sectors only just have the facility to monitor yield, the grains industry is now able to take the next step and map quality. Sydney-based company NIR Technologies has now launched its whole grain analyzer which measures and maps protein (wheat and barley) or oil content (canola) and moisture in real time.

Soil And Water

While automated nutrient soil mapping remains a missing link in precision farming, Australian farmers are putting maps of soil properties to a range of uses.

For example, Western Australian grain grower Simon Wallwork has used a combination of gamma-radiometric and electromagnetic (EM) soil maps to identify his two key soil types — sand and gravel. After “ground truthing” the maps with soil cores, he used them to locate on-farm trials to address questions relating to the choice of machinery and stubble (crop residue) or subsoil treatments. For example, he compared a DBS no-till tyne seeding system with a disc on his two key soil types. He found the tyne resulted in about 750 kg per hectare yield improvement across all soil types.

Others are making profitable use of on-the-go pH sampling for variable-rate lime applications.

Electromagnetic soil maps plus topography data are being used to produce farm layouts designs for irrigation and controlled traffic farming and for the location of soil moisture probes.

Such data is being used by Australian systems (Optisurface and TerraCutter) in providing commercial land grading systems that use GPS and precision technologies. These systems produce a land forming plan that capitalizes on any natural fall in a field. Consequently they can dramatically reduce the volume of topsoil moved, compared to laser systems, resulting in less damage to soil and a quicker payback.

In cotton, canopy temperature gathered from a thermal sensor attached to a low flying aircraft is being tested as another tool to improve irrigation management. Canopy temperature relates to the level of evapo-transpiration by the plant, so is being used as a surrogate measurement of plant available water. Plants under water stress are hotter and those plants often tend to be less vigorous.

Researchers at CSIRO Land and Water are combining vegetation indexes and weather forecasts to produce seven day forecasts of crop water requirements. Indexes are calculated from visible and near infrared data to provide site specific crop water use data on a 30 X 30 m pixel using free spatial data sources and GIS tools. Commercial irrigators are currently testing the system.

Varying Inputs

Spatially varying input rates tends to be limited to more innovative farmers.

In the grains industry seeding phosphorus based on last year’s crop removal plus a uniform base rate is the most commonly varied input.

Variable-rate nitrogen, based on in-crop biomass/chlorophyll sensing is becoming more popular in the higher rainfall or more reliable rainfall grain production regions. Varying seed and fertiliser rate to soil type is become more widely adopted in the lower rainfall Mallee regions where there are distinct bands of sandy and heavier soils.

Perhaps one of the most exciting developments in variable-rate inputs coming out of Australia is the Dynamic Aerial Survey algorithm developed by Dr. Greg Falzon, the University of New England, New South Wales. This can predict and control nitrogen rates on-the-go from aerial applications. Nitrogen rate is calculated based on the area of crop over which the plane has just flown. Rates can be changed every 328 feet when flying at 115 miles per hour.

Weed sensing to reduce herbicide use during the summer fallow period has proved extremely profitable in low rainfall regions.


Autonomous mining trucks already run around Australian goldfields, so it is hoped that autonomous farm machinery will soon be here. Six research centres are working on autonomous equipment including unmanned aerial vehicles.

Machine vision-based weed detection systems have been developed for the sugar, cotton and pyrethrum industries by researchers at the National Centre for Engineering in Agriculture, Queensland The systems analyze color and depth, enabling them to deal with more complex canopies much faster and can identify weeds under crop plants.

The “Ladybird” is the latest robot from the University of Sydney team who also developed the first autonomous seeder for grain crops. Designed and built specifically for the vegetable industry the “Ladybird” is a solar powered ground robot that could be mistaken for a small glasshouse moving across the field. It has an array of sensors for detecting vegetable growth and pest species and has a robotic arm for removing weeds or autonomous harvesting.

Queensland farmer Andrew Bates is helping drive the development of robots suited to large scale grain production activities. His business SwarmFarm has been established to explore, develop and commercialize the capabilities of autonomous farm machinery.

His aim is for each SwarmFarm robots to be small and light, extremely simple with electric drives, very few moving parts and matched to modular components that can be quickly swapped in the field.

SwarmFarm is a project partner with the University of Sydney and Queensland University of Technology in “Robots in zero-tillage agriculture” and his 10,000 acre property is now the testing ground for the small scale robots produced by the project.

Leave a Reply