Marije Krechting works in Asset and Product Management at Alliander. She is the product owner of a DevOps/Scrum team. “We work for the maintenance department's electrical assets. Questions we deal with include: how to perform smart maintenance? How do you make use of all the available inspection data? We used to do time-based maintenance. Once every so many years, we inspected all the assets. For example, we have 12,000 protection relays that we checked every six years. That results in an average of 2,000 every year.”
Questions we deal with include: how to perform smart maintenance? How do you make use of all the available inspection data?
Marije Krechting (Asset and Product Management Alliander)
Smart maintenance assignment
“We started analysing all that data. At that point we noticed that when inspected, some relays had a higher defect risk than others. There had to be a much smarter way to cope with that.” Marije continues. “For about two years now, we have been doing just that, with a data-driven (smart) maintenance order. This allows us to focus on the high-risk assets in the power grid, thus reducing maintenance and using mechanics to expand the grid. We use SAP Analytics Cloud (SAC) to show the analyses, which we call views. SAC contains all assets and all views. Very important: we also use it to provide insight into data quality.”
Choice of SAC
“We switched from SAP Business Objects (BO), a Business Intelligence (BI) platform, to SAC for planning because of HANA's good performance,” Marije continues. “Another reason was that the system also contains many tools. What's more, SAC is future-proof. Erik figured out for us how we could make the switch to SAC: including blueprints and templates for the stories and especially the apps.”
Modifying data in SAC
Noteworthy is that Alliander not only uses SAC for analyses, but also for maintenance planning. Marije: “We make an analysis of maintenance that has already happened. Based on that, we make an annual assignment of the assets that need maintenance in the coming year. This is done using policy rules programmed in HANA SQL. These rules provide us with a prioritisation: which assets should be tackled first? We discovered that so much more can be done with SAC than just making a dashboard and analysis.”
“We do not use the standard tools that SAC planning provides,” Erik stresses. “We have created our own planning in which we can give priority to certain policy rules. That way, we get an overview of assets that require maintenance. And based on that, a proposal comes up for approval.” Marije: “It could be that, according to the policy rules, maintenance is needed in, say, two hundred spaces in a certain region. Then that region may indicate that they only have hours and mechanics for a hundred spaces. Which hundred spaces will you check then? That too is included in our policy rules. The spaces that pose the highest risk, that have not been maintained for the longest time, or that are of a certain type, then move to the top of the list. Those are then maintained first.”
Automated priority list
“The whole process is automated,” Marije continues. “The priority list is created using business rules. You can specify per region how many you want to maintain. Based on those rules, the respective lists are created. We can still make adjustments to the maintenance list before it is issued; individual spaces can still be removed, and rules can be adjusted. An example of these adjustments is the number of assets you want to maintain per region. However, you cannot modify the policy rules themselves! This is of course the intention by design.” Once the list is final, all it requires is the 'push of a button' to send it to SAP via an API. It is at this point that actual maintenance orders are created.
Analysis in dashboards and SAC apps
“The analysis is all done in the dashboards and SAC apps. But adjusting numbers, which they can do by region, is done through UI5,” Erik clarifies. “This is opened through SAC, which all happens outside the view of the user. Subsequently, the logic goes through HANA as backend and is included somewhere in the procedures. Approval and similar tasks are all little UI5 apps displayed in SAC. Those trigger all kinds of database procedures in the background. In turn, these procedures then become the maintenance orders. In the future, those will be entered directly in SAP Quality Management (QM) as Customers Service Orders. Because the orders go back to SAP, we can monitor how we are doing. The strength of this is that for the user, everything is again just in SAC. So they don't have to look elsewhere as well. They can use SAC for analysis, for planning and for collaboration. They give an order and can still look at it together, whether this is what is needed? The fact that it is UI5 does not matter to the user. This approach is quite similar to Predictive Maintenance. That is also where we want to go.”
Improvements in the future
Hence, a lot can already be done with SAC, according to Marijke and Erik, but some things are still under development at SAP: “A visual representation of geo-plots is not yet possible, for example. The same goes for including photos. SAP is probably still working on that, as it was possible in the old system. Another example is filtering into regions. This is possible in Stories but not yet in apps. For instance, you cannot yet filter out all cities in the Netherlands. We have developed that ourselves. It is great that this is possible, that SAP gives us that freedom.
Furthermore, we would like to see a combination of a dashboard with write-back functionality to be able to manipulate data. You actually want to do that in those control tables. Obviously, you don't want to sit on the source data. The large amount of data we have is very valuable. In the future, we might start analysing it with Artificial Intelligence (AI). We also want to start recording expert knowledge so that we get a learning system with policy rules.”
Alliander develops and operates energy networks. Through our cables and pipes, over three million Dutch households and companies are supplied with electricity, gas and heating. They operate a 90,000km electricity grid and a 40,000km gas network, and take great pride in our networks being among the world’s most reliable. Their 7,000 colleagues make sure the lights are on, homes are heated and businesses can keep operating.
This article was written and published by the VNSG. The VNSG is the Association of Dutch-speaking SAP Users in The Netherlands.
A special thanks to Marije Krechting (Alliander) and Dirk-Jan Schenk (Editorial board VNSG).
Photo: Erik Leemans from Expertum gives a presentation at the VNSG