Most rural water businesses spend a lot of money on data collection. Some of it is used for operations and is considered useful only at the time it is collected since it describes the current situation. Some is collected to fulfil reporting obligations, say to the meteorological authorities or environmental protection authorities and is not considered useful beyond fulfilling those obligations.

Taking a manual water quality sample in the Werribee River, Australia

Some data is collected by field visits, say bore readings or water quality samples, while others are provided by telemetry (with costs of equipment, communications and maintenance). Some comes from third parties with a contract to supply the data.

Just consider the range of date: bore levels, water quality, salinity, weather stations, maintenance work orders, customer details, customer orders, meter readings, allocation and share trades, storage, channel or river levels and flows, scada alarms, soil moisture readings, soil survey data. The list goes on.

How much data is collected, is astounding. Take the example of a single automated regulator gate regularly generating flow and level information in a channel as well as other data like battery charge, solar panel charge rate, battery temperature and so on. Let’s say it sends 4 kilobytes of data every minute during the irrigation season which runs for 8 months. That adds up to 1.3 Gigabytes of data. If you had 800 such gates, you’ll generate a Terabyte of data in one year. Often, there is so much data, that it is archived and stored where it never sees the light of day again.

Regulator gates generate enormous amounts of data. Photo: Michael Kai.

Incredibly, given the costs of collection, few rural water businesses spend time analysing all this data. Sure, it gets used for its intended purpose, say a meter reading to produce an invoice. But there is remarkable value there waiting to be discovered through analysis.  In the words of Michael Dell in the Irish Examiner, “If you look at companies today, most of them are not very good at using the data they have to make better decisions in real time”.

Imagine that you could compare years of data to see trends and patterns. For example, to extract every water order made by a particular farmer over the last 10 years to demonstrate the improved efficiency resulting from modernisation. Or gain an insight into the impact of a high rainfall event by assessing the impact of a similar event several years ago. Imagine being able to show improved service levels in a channel after it was weed-treated. Or finding which channel subsystems have had the worst water balances or service levels in the last five years so you can target maintenance spend.

miops

The data you are already collecting has the power to reduce asset maintenance costs, better target investment in modernisation, improve water harvesting and delivery efficiency.

Can you afford not to analyse it?