While machines are generating more data with increasing granularity, our most seasoned experts are leaving with all their accumulated experience. So why are we buried in data and starving for insight?

Common sense says once every sensor, machine and system is integrated, we can start to reap the benefits. But what if we’re looking at the problem from the wrong end of the telescope? David Bartolo, Head of Asset Intelligence, AGL says too often we’re ignoring the very methods humans use to build their own expertise, and disruptive technologies allow us to harness that, and exploit what we already have to drive new levels of performance. He recently presented at Mainstream OnAir 2021 and shared how organisations can use what they already have to drive lasting transformation. Here's a summary of his session.‍

The Future of AGL – Using What You Have

‍The energy industry in Australia is going through a massive transition, and AGL is right at the centre of it. AGL is in the process of splitting most of the assets into a new company which will be much more customer-centric. Having worked in the energy industry for a long time, David is aware of the need to concentrate on the now, while also planning for the future of the power generation side of the business and utilities. “We’re going to have to become a very low-cost provider of centralised energy for our customer base in this new business. And we’re going to have to use every trick in the book, every technology, every process to do that,” David said. Along with being safe, reliable, and low-cost, the company must also maintain a positive cash-flow. David concedes this won’t be easy, since they’re competing with a flood of renewable energy coming in for free. “We must transform this new business from a fossil-fuel-based business to a renewable-based business in a relatively short space of time,” he says. David believes technology is going to play a pivotal role in achieving their strategic goals. “We’re going to need thousands upon thousands of IoT sensors added to our processes, so we have more situational awareness of what the assets are going through and their condition.” David adds “We’re going to have to use that data to shift maintenance from periodic-preventative, into true condition-based maintenance, because there won’t be money to do anything else.” However, it’s not about getting rid of the old and replacing it with something entirely new. David explains that AGL have already done a lot of useful work in that space by gathering data which is both time-series and condition-based. “Getting that firing on all 8 cylinders and getting the human disciplines to give us quality data – and then, in the future, combining those two data sets to really help us optimise every dollar spent on maintenance. I think that’s the challenge for us in the future,” David observed.

‍AGL’s Vision for Transformation

‍David admitted in the beginning it was all quite confronting because there was no vision of the future back then. But they knew they needed something. “We were completely blind,” David said. “It was quite embarrassing asking for print outs of trends to try to work out what had happened on a plant,” he said. The original system wasn’t centrally accessible, and had issues with how it could be visualised and shared. It could only be consumed by very few and on a limited scale – meaning you could only get limited value from it. At the time, a key solution was implemented to treat just one problem, but instead it set AGL up for success for the next 7-8 years. That solution was to invest in the best time-series data infrastructure AGL could get for their money. “As part of the investment in the infrastructure, we didn’t want that to only be used by a small number of people because that’s not actually empowering the corporation with data. So as soon as it was in place, we trained up over 450 people to use that data infrastructure. By doing this we’ve empowered everyday engineers, everyday technicians, and everyday managers to consume the data and try to harness value from it,” David explained.

‍Advances In Technology Enabled Greater Transformation

‍David soon realised that other things became easier as technology changed. “We could do analytics, we could do diagnostics using really fancy software that’s easy to use, and we could do it on scale because we had the data from every site,” he reported. Being able to combine and compare data sets generated other insights, including better decision making. When the wave of IoT technology came along, AGL could afford to install sensors in the field because the technology was no longer cost-prohibitive. “That’s a game changer. We had a home for that data which we could bring in securely, easily, and it could be correlated with everything else that was happening on the plant. It made the data much more valuable, rather than being isolated somewhere,” David explained. AGL had a vision of empowering the corporation with this data and then harmoniously introducing new technologies into what they call ‘the operational technology ecosystem’. This included the introduction of electronic logbooks in the control rooms, and the ability to integrate with the historian infrastructure. “This allowed the importation of time-series data to the operator in real time and automation and improvements to shift handovers, which were things we’d never even dreamed of,” David admitted.

‍Important Lessons Learned During the Transformation

‍By engaging with other businesses who had experienced transformations of this nature, AGL learned that one issue is the volume of data crunching, rather than the complexity of the process. The data generated should allow innovation, analysis, and the ability to gain valuable insights quickly.David believes the way to achieve this is to have someone who has a good understanding of the technical processes first, before applying the data science. Having someone who understands the data landscape can apply the analytics easily, and at low cost. David shared a few parting words on making transformation work:

Don’t:

• Let people loose on the data who don’t have a good understanding of the process. Skills first, and align those skills with some data science, or at least put those people together very closely.

Do:

• Centralise the data and empower your people to use it. Just like they use Excel every day, they should be able to use that data infrastructure every day. That delivers good outcomes, and good engagement with your people as well.
• For heavy assets, consider putting in diagnostics as soon as you can, and put some dedicated people on it. If they’re out on the plant and they’re preparing for the next outage, the diagnostic system won’t work. You need to be dedicated to those alarms, filtering out the garbage and then providing quality insights to those people to help avoid loss.