Why utilities need to refine their approach to data management in 2020
Here’s what the landscape for utility companies in 2020 looks like:
Climate change exacerbates extreme weather conditions like wildfires. These disasters threaten power companies’ infrastructure, like transmission and distribution lines, poles and towers, the things that customers rely on for power. As a result, outages become more frequent – either because the infrastructure is damaged, or because utilities are pre-emptively shutting the power off to prevent, for instance, flowing electricity from a downed power line from potentially causing more harm. In either case, safety for both the public and company personnel drops. Customer dissatisfaction grows louder. The costs of maintaining the status quo, and making basic infrastructure repairs and maintenance, continue to become more expensive.
And to top it all off, the ability to foresee and manage these disasters – which affect everything from asset management and infrastructure integrity to environmental risks and consumer trust – is being constantly overwhelmed by a volume of inspection data that is simply too great for human beings to handle alone. What good are inspections if there’s simply too much data to process, organise and leverage into something actionable?
It’s a new year and a new decade. With that comes a new opportunity for utility companies to redefine and modernise how they are overseeing their infrastructure assets. The path forward lies in collecting and processing inspection data via a holistic data approach driven by AI and inspection automation.
Building a circle of inspection automation:
Power companies have a specific set of inspection needs. They plan out mission flight paths with different vehicles (drones, helicopters or fixed-wing aircraft), armed with various types of sensors – colour (RGB), thermal, hyperspectral, LiDAR or some combination of the four – to capture different sets of data in the field. These drones or aircraft proceed along their flight paths, collecting information on everything from the physical status of transmission lines and poles, to environmental and weather damage, to the proximity of vegetation that, if near a live wire, could spark a fire.
But these flights produce thousands upon thousands of images, up to petabytes of raw data that is simply too laborious and time-consuming for human beings to sift through, let alone draw conclusions from. This is the heart of the need for inspection automation, and why utility companies will have to implement AI- and machine learning-based automation tools into their data workflows.
These AI and ML-driven approaches provide fast, deep data inspection and analysis, scanning images more rapidly than any human can. This is all to establish patterns across data sets, create baselines and context, and organise data into more intuitive and actionable workflows. Data collection is only useful to the extent that the information being collected can actually be understood and acted on. Now that we’ve reached a tipping point where there’s more data than any one person could manage, AI and ML have become essential ingredients for utility companies’ data management practices.
By deriving these more useful, constructive conclusions out of more easily managed data, utilities can then readjust their inspection needs and processes accordingly based on what they need to evaluate on the next flight path – starting the full circle of this inspection automation process all over again.
A holistic approach:
Equally important to the automation piece is the necessity for adopting a holistic approach to data collection and management. IT organisations in general – and this is true of many industries, not just utility companies – have historically suffered from a data silo problem. That means different data sets, generated from different sources, stored across disparate databases with little if any overlap or communication between the teams owning them. As a result, the organisation as a whole ends up accumulating redundant levels of data or taking actions that fail to take the big picture into account – because the decisionmakers can’t see that bigger picture.
By taking a more holistic approach to how inspection data is being collected and processed, utilities can ensure from the start that all information is being pooled together – regardless of the source of the data or the team handling it. Everything is consolidated. This has two immediate benefits. For one, consolidated and organised data makes the information much more easily available and searchable for future reference.
Additionally, this holistic approach to storing and analysing data generates workflows and actions that take all available information into account, not just whatever a specific team or individual happens to have access to. This creates a more informed and comprehensive approach to inspecting utility infrastructure assets, compared to the piecemeal system that’s built around using different tools with separate, sometimes conflicting, ways of gathering, storing and processing inspection information.
Achieving three key goals via inspection automation:
Utilities are facing a wide array of data management challenges: information overload, human error, a lack of objective decision-making tools, a lack of efficient coordination between teams in the workplace, slow data collection and lengthy data analysis processes. All of these problems are compounded by a status quo of utilising disparate data management tools that simply don’t play well together.
Adopting a holistic data approach, built around a circle of inspection automation, eliminates the pitfalls of multiple siloed systems and streamlines the data collection process. One pool of data, organised and prioritised into more intuitive workflows, provides the insights and actionable next steps that utility companies need in order to reduce costs, increase safety (for both the public and their own personnel) and make better-informed decisions. At the end of the day, these are the three goals that all utility companies are striving for in their data management practices – and a system of holistic data analysis and inspection automation ticks off all three boxes.
Ilkka Hiidenheimo, CEO, Sharper Shape