On-farm experimentation

TractorProject vision: to help farmers tailor paddock-by-paddock management through on-farm experimentation and analytics

Digital agriculture is certainly a hot topic. A range of new data streams are becoming available but can they all be turned into actionable information that add value to farming operations?

While there are significant commercial developments around using this data for benchmarking and monitoring, there are surprisingly few systems that support on-farm analysis and planning. In particular, there is a lack of technology that integrates information sources across the whole-of-farm enterprise scale, with design, implementation and evaluation of farming plans.

Digiscape’s on-farm experimentation project addresses this opportunity.

Our aim is to develop a prototype software system that will represent different spatial scales (for example plots, blocks or paddocks), different farmer actions (variety planted and management options, for example fertiliser applied) and the different machinery used.

The system will link external data sources to these visual interfaces for farm planning and analytics. Farmers will be able to use the software to explore existing data and refine farm management questions, plan an operational farm-scale experiment, implement that experiment, and analyse the outcomes. Once completed, the data and analytics will be fed back into the software system and be available for subsequent on-farm planning. These experiments will typically involve a small number of treatments (such as variety x management options) and be simpler than complex field trials.

Farmer with tractor and ipadAn example experiment could be to compare yield across a winery for two grape varieties under two different mulching strategies.

It might sound simple, but nevertheless, it’s a bold aim. Coupled with deep understanding of the needs of specific agricultural sectors, it requires engineering, experimental design and analytics skills. There are scientific and engineering challenges to make it intuitive and efficient.

However, it will be applicable across both broad scale and intensive agriculture, both crops and livestock and it has the potential to be used throughout the world’s varied production systems for real-world agricultural strategic planning.

This project is led by Dr Brent Henderson.