Digifarm robot


Agriculture technology (AgTech) for improved production and environment

We're supporting agribusiness by developing a digitally enabled network, which will simultaneously monitor crop production, animal production, and soil & ecosystem health.

DigiFarm overview

Watch an overview of the DigiFarm project.

DigiFarm (2019-2023), funded by the National Landcare Program, brought together the community, farmers and environmental stakeholders as part of an integrated approach by the University of Sydney to its farming, education and research activities.

The DigiFarm project aimed to develop a digitally enabled network to simultaneously monitor crop, soil, and animal production as well as biodiversity and ecosystem health including native flora and fauna. This network aimed to enable the triple bottom line framework of social, environmental and financial accounting to optimally manage a production ecosystem.

The project aimed to transform Llara Farm into a digital agile, production and environmentally resilient ‘digi-ecofarm’ of the future which is a beacon for the north-west NSW region. It has done this through the building of a physical and virtual DigiFarm hub at Llara in addition to a network of ten satellite farms across north-west NSW which provide digital dashboards of health, production and social metrics. This aimed to create an education platform for farmers, agribusiness, schools and other stakeholders to experience the latest ag-innovation thinking.

The project involved 20 collaborators and has demonstrated a range of digital technologies to farmers across the north-west NSW region through 13 satellite demonstration sites across North-West NSW. It has delivered outcomes from a multidisciplinary perspective relating to cropping, livestock, ecology, the environment, and agricultural technology. The project has highlighted the benefits of integrated sensing and is an excellent showcase for digital agriculture for the agricultural and environmental sectors, the government and community.

Aerial view of Llara Farm, Narrabri

Aerial view of Llara Farm, Narrabri

Check out all the action from the 2022 Digifarm Expo webinar series

 2022 DigiFarm Expo Webinar Series Session 1 - Cloud to paddock

 2022 DigiFarm Expo Webinar Series Session 2 - Cropping, the good and the bad


2022 DigiFarm Expo Webinar Series Session 3 - Livestock



2022 DigiFarm Expo Webinar Series Session Four - Farm to International Fashion Market

Digifarm in detail

The benefit of digitising Australian agriculture was estimated at $20.3bn annually.  Agriculture was also said to lag other industries in digital adoption (see https://www.crdc.com.au/precision-to-decision). Landcare administered grants for ‘smart farming partnerships’ in an attempt to evaluate technologies for farm readiness and to improve adoption. From this, the University of Sydney’s Digifarm at Narrabri was born.

Digifarm was a collection of subprojects that combined research with the Landcare mission of farmer extension. By utilising both extension and research, we evaluated whether technologies are suitable now, and also whether they are likely to remain suitable in the future.

The University of Sydney has been in Narrabri for 60 years, operating a distinguished wheat breeding program. It has a 2600ha property, Llara, used for cropping and beef production. It also runs around 25,000 crop trial plots per year and contains remnant native forest.

The DigiFarm project completely changed the way in which spatial, farm management and sensor data are managed and visualised on the University of Sydney’s Farms. The use of AgWorld (cropping) and AgriWebb (livestock) commercial platforms, integrated with precision agriculture software such as PCT Ag (crop spatial information) and Cibo Labs (pasture spatial information), has allowed data to be managed in a much more efficient way. This allows better informed management decisions to be made as well as ensuring full use and benefit of data obtained on the farm.

The DigiFarm project has led to the establishment of the Pairtree dashboard which is a holistic dashboard displaying integrated information from all of the other digital platforms and services being used. By bringing together the data and visualisations from all platforms and digital technologies (E.g. soil capacitance probes) on the farm, the dashboard allows efficient and effective farm management decisions to be made.

Farms comprise more than 50% of the Australian land surface, and a much higher proportion of specific habitats.  Farmers are therefore custodians of much of Australia’s wildlife. We installed and monitored a grid of 20 camera, 20 audio, and 12 bat-echo traps across the farm, to quantify how wildlife are using the different farm landscapes. The camera traps captured 55 species from 1 million wildlife images which were captured. The bat acoustic sensors monitored the echolocation calls of bats for 3.5 years, capturing 440,944 bat calls which allowed the identification of 14 unique species, 5 of which are threatened. Over the same period, the bird acoustic recorders identified 140 unique bird species on Llara. Alongside acoustic recorders and cameras, we also conducted human surveys of birds, flora, and invertebrates and installed smart traps for mirids, a pest of cotton, which evaluate their numbers on a daily basis. Over the course of the project, 100,000 invertebrate individuals were counted with 107 species groups identified.

In the future, it is predicted that landowners will be able to buy a sensor at an affordable price, place it somewhere on their land, and receive regular updates of the number and diversity of birds and bats simply by email.  Making this a reality required the training of AI algorithms to recognise calls and extract meaningful information. These AI models were used in the DigiFarm project to identify the species associated with the acoustic recordings. The data gathered from this aspect of the DigiFarm project has contributed to building a database of the natural capital on Llara along with successfully developing and validating AI models to recognise and extract bird and bat calls. 

Soil moisture models for dryland cropping can improve predictions of future yield, and thereby help farmers to spend the right amount of money on inputs (seed, fertiliser) to optimise profits. Good models combine big data from the Bureau of Meteorology and satellites, to local data from on-farm soil moisture probes and rain gauges. 

Fieldwork, lab analysis and modelling were used to characterise the soils of Llara in terms of plant available water capacity (PAWC) and chemical soil constraints. Mapping of soil chemical constraints was carried out and digital soil maps of depth to constraint were produced for pH, salinity, sodicity and depth to first encountered constraint.

Spatial nowcasting of plant available water (PAW) was implemented using the water balance modelling approach which was used to predict and map the soil water content for the top 1 metre of soil across Llara with and without constraints considered. This spatially distributed soil water balance model can now be used to nowcast plant available soil water. This improves our knowledge and understanding of soil moisture and field constraints to allow for better and more effective management, resulting in enhanced productivity and natural resource management (NRM) outcomes. 

Four networks were installed and trialled on Llara Farm as part of the DigiFarm project to allow for 24/7 remote and automated farm sensing using equipment such as soil moisture probes and water tank monitors. A range of digital equipment from 4 companies (Goanna Ag, EnviroNode, ICT International and Hussat Probes) was then trialled and tested to evaluate its performance and reliability.

56 digital soil moisture probes were installed, with the intention of (1) improving predictive models, (2) evaluating the right number of soil moisture probes needed in dryland systems and (3) testing the networks which we installed. Hence, we have evaluated sensors from different suppliers, found some of the practical issues of maintaining them on an active farm, and have developed simulations that will find their way into yield prediction applications.

The output data of real-time soil moisture and soil temperature from the probes was then displayed on the Pairtree dashboard, allowing efficient and effective farm management decisions to be made. These probes made use of the 4 networks which were installed across the farm to allow for 24/7 remote and automated farm sending and monitoring. In addition to soil moisture probes, we also installed and tested water trough sensors, 5 water tank sensors, tipping bucket automated rain gauges and 3 automatic digital weather stations on the networks we installed. 

Drones, satellites and ground data were used to improve the APSIM plant growth model of wheat under different levels of nitrogen. The findings of this sub-project of DigiFarm indicated that crop condition and production simulation in the APSIM crop model can be improved by assimilating remote and proximal sensing observations, leading to improved yield prediction.

Precision weed management is developing rapidly, with emerging AI algorithms and matching hardware. Commercial products are available for green-on-brown (green weeds against bare soil) but green-on-green (weeds within crops) is harder. The process involves identifying a weed in real-time as the tractor (or field robot) passes and performing an action. 

An Agerris ‘Digital Farmhand’ (DFH) robot was purchased and tested for its precision weed management capabilities as part of the project. A digital camera-based weed detection system and accompanying spot-spraying system were developed and mounted on the Agerris Digital Farm Hand robot to enable site-specific fallow and fence line weed control. The robot and incorporated system which was made is fully capable of autonomous spot-spraying weed control in fallows and along fence lines. This technology was showcased and demonstrated at a number of extension events.

In addition to this, as part of this project activity, we also developed two Open Spot Sprayer (OSS) units and produced a how-to for a low cost, camera-based fallow sprayer module. The 3D printed housing model files, detection code and instructions for construction were made as freely available as possible. The module is boom mountable and demonstrate the current capability of low-cost embedded computers, with a demonstration model developed for use at the field day. This technology was then used at two LLS demonstration sites across Central and Northern NSW. Following the success of the autonomous spot sprayer unit, the Open Weed Locator (OWL) platform was developed to highlight the readily accessible, low-cost opportunity for fallow weed detection. This platform will play a vital hands-on educational role for the use of imagebased weed detection, providing an exciting opportunity for community involvement and the future development of the technology.

The DigiFarm project tested Natural Sequence Farming (NSF) water management techniques of contours and the installation of a combination of rocks, logs and vegetation in a creek to spread the water across the pasture landscape. NSF is a regenerative practice that seeks to restore landscape function by re-establishing the ‘chain of ponds’ within Australian watercourses. Along with managing vegetation and grazing, slowing the flow of water with the formation of contours on open landscapes or constructing leaky weirs in creeks and gullies can re-establish natural hydrological functions in the landscape, riparian zones and surrounding floodplains.

We aimed to answer questions around the effectiveness of these activities for landscape rehydration and soil repair, enhanced biodiversity, maintaining higher soil moisture, generating more pasture, building carbon stocks and improving water quality. Measurements taken included groundwater quantification via in situ real time probes and neutron probes, soil carbon, nitrogen and bulk density at depth, soil decomposition and soil organic matter (SOM) formation and biomass production and quality, among others.

Cover cropping trials have also been carried out as part of the project with mixed species cover crops incorporating various legume, brassica, and flowering species into millet and oat base crops. Moisture monitoring identified that in cover cropped plots, rainwater harvesting and crop water extraction respectively increased by 50% and 25% compared with a fallow control. Importantly though for the region’s dryland growers, moisture deficits incurred by cover crop establishment were replenished by the time wheat was sown. Soil monitoring identified significant improvements in stability, particularly for carbon and structure. Results for the latter, at cover crop termination demonstrate that aggregate disintegration due to water ingress was more than halved for cover cropped plots. These trials aimed to better understand the benefits that cover crops can bring to the farming system with more expansive trials planned for 2023-2026 which will monitor crop, soil, water, biology and financial system attributes. 

The DigiFarm project also trialled native grains with the aim of finding a suitable balance between production and natural ecosystems. This was done by evaluating the triple bottom line metrics of including a native grain enterprise on a farm while also bringing together Indigenous and non-Indigenous people in pursuit of this goal.

This aspect of the DigiFarm project established connections between growers/landholders and the Indigenous community for respectful knowledge sharing and capacity building for Indigenous leadership of the native grains industry as it emerges. Methods for seed harvest and processing of native grains were also developed as part of the project. We also used the native grains seed increase area and the ability to produce food-grade native grain (which cannot be purchased commercially) as the catalyst for multiple research projects beyond DigiFarm.

Soil testing usually means sending something to a lab, but we are experimenting with simple tests that farmers can do in the field to quickly assess a soil.  The SLAKES app and the teabag index have been conducted across our farm and compared to prior maps of soil type.

Over the course of the project, over 200 soil tests were performed and analysed. 58 soil health monitoring sites were identified based on combinations of soil type and land use (cropping, pasture, forest) by contrasting seasons. Soil data was collected for chemical, physical, and biological assessments of soil structure via the slaking index and of the soil microbial activity and decomposition patterns across seasons via the Tea Bag Index. 

We worked with livestock production to develop new technologies to better manage livestock and pastures through real-time data for improved decision making. This information is important to optimise livestock production, health and welfare, and sustainability with digital technologies. Several subprojects were conducted using cattle weighing systems, including in-yard walkover weighers, and in-field Optiweighs. The latter is a young Australian company which has produced what is estimated to be the world’s first in-field cattle weighing system. Cattle are attracted to the weigher by a lick block, and are not stressed, whereas they need persuasion to use a walkover weigher. This technology gives the producer a daily estimate of the herd liveweight, growth rate and animal behaviour, allowing them to better plan pasture utilisation and time to sell, improving productivity and efficiency of production.

As part of the DigiFarm project, we carried out a community-level project using GPS collars to track the movements of feral pigs in order to provide valuable insights into how the local population spatially utilise the Mulgate Creek valley throughout the year. We collared 15 pigs and then gathered GPS data from their collars. Analysis of the data showed that the collared pigs utilised different parts of the landscapes throughout different seasons. This is valuable information when targeting the animal as it indicates that throughout the year the animal is utilising different resources allowing an informed decision to be made to target the animal when it is in the most suitable location for effective management.

A farmer looking at crops at Narrabri farm

A farmer looking at crops at Narrabri farm