In an age where digitisation is revolutionizing industries, water utilities and councils are not exempt from the transformative power of data-driven technologies. However, as utilities embark on their digitisation journey, they face a critical challenge: data gluttony. The increasing deployment of solutions and myriad selection of platforms has led to an overwhelming influx of data, creating operational burdens and complicating the business model. Gerhard Loots, IoT expert and CEO of Kallipr, one of Australia’s leading IoT solution providers, gives his perspective.
It’s said that the more things change, the more they stay the same. Since the 80s, we’ve attached telemetry units to assets to manage them more efficiently. As the cost of solutions decreased, the propensity of solution deployment increased. An unintended consequence was that the size of operation rooms had to increase to deal with the vast new amount of data.
One area where this pattern emerged was the rapid popularity of security cameras. The sudden influx of video data overwhelmed operation rooms. A UK study found that the average person could only look at 4 cameras at a time, which, when calculated over a 24/7 operations model, meant you required one person per camera.
The business model needed to be fixed. But to solve the issue, video analytics was introduced. Today we know this as video AI. AI-enabled operators had to look at black screens that only lit up when a camera feed showed something was out of the ordinary. The business model was fixed, and the unintended consequence was mitigated.
The same issues are now seen as water utilities leverage advanced technologies to streamline water utility operations. As more and more data becomes available, it’s integral that we find a way to manage it effectively.
Data needs purpose
The premise is simple: We don’t need to know everything is going as expected. We need to know when it’s progressing away from, or not in, the expected state. The operational process of managing by exception is widely used beyond industrial applications. When you consider the oil light on your vehicle, the light only goes on when something is wrong. You assume it’s right otherwise.
Although telemetry solutions like the Captis product can alert only when there’s an issue, digitisation hasn’t reached this maturity point yet in our experience at Kallipr.
When engaging with customers at the start of their digitisation journey, we consistently encounter data gluttony that complicates the business model and usually isn’t the most efficient way of solving the problem. Dashboards showing everything is OK are useful during the trial or concept phases so the business can learn to trust the data and create baselines for future analysis. However, as the deployment of the devices scales up, operation centres will increasingly be under pressure to provide more people to manage the data.
The utilities that are doing telemetry well are those that are very clear about which data is important. At Kallipr, we encourage our customers to create a baseline of Normal data (green), and abnormal data can be split into Watch (amber) or Act (red). By understanding what green or normal data is, everything-is-ok data can be discarded or stored at the edge, reducing data and cloud consumption whilst also being more environmentally friendly.
As events get refined into amber or red, the percentage of amber events should decrease over time as it’s classified as green or red through machine-learning patterns.
Machine-learning patterns support better ticketing
Machine-learning patterns further decrease the load on operation centres by automating red events that require action through an automatic dispatch ticket rather than manual intervention from an operator. While this may sound intimidating, computer-generated dispatch commands are already part of our present: Uber and Taxi Apps have replaced taxi booking services.
The digitisation journey’s final stage (and highest value) of automation requires clarity on the problem being solved. Data in new places enables us to make better decisions, which can be optimised even further when we anticipate the unintended consequence and solve this for them. People don’t need data; they need insights. And if we can automate these insights into outcomes, we’ll achieve a more sustainable future.
From our experience at Kallipr and working with water utilities at all stages of their digitisation journey, the following considerations can help to achieve this faster:
- Organisational alignment on data framework and policy
- Operational clarity on green state and red state data
- Organisational commitment towards field service automation
- Pragmatic digitisation programs that grow with the organisation’s acceptance of data-driven decision automation
Ultimately, the utility of the future will manage this influx of data by being very clear about which data is important – harnessing insights from the data rather than drowning in its sheer volume. By effectively managing data, utilities can optimise their operations, achieve greater efficiency, and pave the way for a more sustainable and digitally empowered future.
For more information, visit https://kallipr.com/
Related Articles:
- CSIRO leader encouraging a positive culture
- A watershed moment for Aus water management
- The power of positive disruption