Agriculture is the ideal target for the development of new big data applications which can convert decades of research into usable information for farmers, according to chief executive of the Cooperative Research Centre for Sheep Industry Innovation (Sheep CRC), James Rowe.
Speaking at yesterday’s Connect Expo Big Data Strategy Summit in Melbourne, Prof. Rowe told the conference that agriculture was particularly well placed for the big data revolution due to its mixture of complex and dynamic characteristics where farmers have to manage soil, plant, animal, climate and market interactions, which all change every day.
“There is already a rich mine of data available from decades of institutional research which has not been fully exploited, including detailed bio-physical models and a number of large data resources that are suitable for ‘re-processing’ using technologies such as machine learning,” Prof. Rowe said.
“The focus has now shifted from dealing with the cost and difficulty in collecting data, to the problem of how to analyse and interpret large amounts of data, and the opportunity to make the information available in a meaningful way for end users via apps.”
The Sheep CRC has made good progress in developing big data products, last year launching the popular RamSelect.com.au app, which helps sheep producers find the right rams for their flock and their production system - quickly, easily, accurately and with confidence that the information presented is supported by large amounts of objective data.
“In the Sheep CRC we are now focussing on developing a second web-based app that combines big-data capabilities with deep subject knowledge, sophisticated bio-physical models and machine learning,” Prof. Rowe said.
“These web-based apps are being designed to make it very easy for sheep producers to make good management decisions based on a large amount of complex data via a simple interface on their phone, computer or tablet.
“The new app we are working on analyses environmental risks, such as weather, nutrition, worms, and flies, as well as the ’susceptibility’ of individual animals within the flock - based on weight, health status, genetic history etc., to make better management decisions about wellbeing and productivity.
“This app will underpin improvements in animal wellbeing and productivity through pro-active management decisions that target reduced on-farm sheep mortality and the selection of more productive animals.”
Prof. Rowe said the rate of data empowerment was increasing exponentially, with easier data gathering, cheaper data storage, faster computing and smarter analyses all converging to deliver very useful information for end users.
“A good example of lower prices is the cost of storing data which has fallen to around 3c/GB compared to $450,000/GB in the early 1980s. Rapidly changing on the equipment side is the adoption of powerful hand-held devices – smartphones and tablets – where sales have gone through the roof in the last three years,” he said.
“The rates of changes taking place in computing speed, cloud computing and mobile device development makes data science a very specialised area, and it is essential for an organisation like the Sheep CRC to team up with experts in the data science area such as the Data-to-Decisions CRC, Telstra and specialist app developers like Pivotal Labs.
“We are also working with farm software companies to link in with the data gathering technology on farm, based on electronic identification linked to measurement systems, which can feed in new information to these big data tools, and allow producers to individualise the analysis they receive for their business.”
More information on the Sheep CRC is available at www.sheepcrc.org.au
Media contact: Michael Thomson on 07 4927 0805 / 0408 819 666