New data service helps farmers

By ADAM GIFFORD

A three-person Te Awamutu farm consultancy has teamed up with Hamilton-based database company Reel Two to create an online information collection system for farmers which will eventually form the basis for data mining.

AgriNetworks managing director Chris Mulvaney said the AgriGain system meant the firm would be able to service thousands of farms, rather than the 100 it advised now.

The company gets its sheep and beef farmer clients to collect data on the physical condition of their animals, soil and climate conditions and financial records.

"By analysing this we can identify factors limiting their business. So, for example, if there is a pasture problem we can refer them to other agribusiness organisations with particular expertise there," Mulvaney said.

Switching from paper to electronic data collection would reduce the time farmers spent waiting for advice.

"By entering their data in the site themselves, the farmers can get an updated report almost immediately, instead of having to wait three or four days for us to get back to them," he said.

AgriNetworks found Reel Two after Technology New Zealand turned down an earlier funding proposal involving a multinational software development firm.

"They would have been very expensive, and I think in the end we got a better result," Mulvaney said.

Technology New Zealand put $195,000 in matching funds into the AgriGain software development.

Reel Two managing director Stuart Inglis said AgriGain was written in Java Server Pages on a MySQL database, and would be hosted by Reel Two at the new Waikato Innovation Park in Ruakura, or "silicon paddock".

"We took AgriNetworks' spreadsheets and, in effect, distilled the expert information from current employees and put it in the system," Inglis said. "Because we used open-source software, we don't have to pay another $100,000 or more for software licences before we started."

The website, which will go live this month, is just the first stage, implementing AgriNetworks' existing system on a web-site, with a little machine learning and some data analysis tools.

It incorporates Weka (Waikato environment for knowledge analysis), an open-source data analysis tool developed at Waikato University.