Facility maintenance is rapidly evolving. How does data impact the future of it?

Disrupting the status quo with data

Typically, maintenance is based on scheduled intervals of checks, inspections and repairs. This approach is labor intensive and only provides an intermittent insight into a building’s status.

In recent years the amount of data available for cloud analytics has increased dramatically.

When I joined Siemens 6 years ago our cloud software typically captured 10-50 data points per building, mainly energy meters. Today we talk about a range of 1’000 – and up to 10’000 – data points per building. This really calls for leveraging the data for maintenance!

Traditionally, inspections and checks are performed to ensure high availability of assets and to avoid unplanned failures. Today, through data-driven services, we can provide savings on the repetitive tasks without losing asset availability, nor increasing the risk of failure.

How we’re modernizing building maintenance

By working with our customers, we can create better conditions for all. The first step is typically to review the existing maintenance tasks and identify which can be fully replaced by analytics, and which can be supported or optimized with the help of analytics.

Inspections and checks are often tasks that can be replaced. These tasks do not include replacement of parts, but instead ensure that operations are running smoothly.
These inspections and checks aim to identify, for example, a leak, or the need for calibration, or cleaning, before the issue impacts the performance of the facility or even risks a larger failure. By implementing analytics, instead of inspecting vast amounts of assets, symptoms can be monitored and resources can be directed only to the assets where an issue has been identified.
Although the task is not necessarily the same when replacing inspections with analytics, the outcome is. For example, visual inspections or listening to an asset would typically need a camera or sound sensor. This is, of course, costly to install and automatically process. The same symptoms can be identified from observing the operating data such as higher flow rates, temperatures, or if valves are opening more than is typical. These tasks are very difficult to carry out with just the naked eye. But analytics can provide a more efficient and cost-effective solution.
Tasks that include replacement of parts can also benefit from analytics, for example where certain tasks are performed yearly, or every 6 months, regardless of the real state of an asset. A useful example here is the yearly replacement of an AHU filter. With the help of analytics, we can better time the maintenance activity by monitoring the filter cleanliness. The filter exchange date can be brought forward or pushed to a later date, depending on whether the filters are clean and performing well, or require replacing.

How can analytics be applied to facility maintenance

There is a perception that to apply analytics to a building maintenance system, many new sensors will be required, requiring financial outlay and considerable disruption. However, it is possible, and indeed encouraged, to work with the existing sensoring and data available in the automation system. Usually, a lot can be achieved using the existing sensoring. There is a large set of predefined analytics rules and algorithms that make it easy to start applying the use of analytics. The main requirements are:

Data availability: ideally, the data would be available in a BACnet network, where existing tools can capture data
Access to the customer’s maintenance plans and tasks.

As a second stage, performing a gap analysis of required sensoring can help achieve a higher level of replacement inspections through analytics. To ensure a high ROI on any additional sensors, it is important to focus on the most frequent tasks and those that historically have had the most corrective maintenance required.

Real-world example: Data-driven maintenance

University Properties of Finland (SYK), a company that manages and maintains the buildings of many Finnish universities, introduced a new building maintenance model at Tampere University in just two years. Working with Siemens, the team at SYK co-created a data-driven model for the 81,000 m2 campus that houses 11,600 students and over 2,000 employees. The application of asset performance services and analytics resulted in numerous benefits:

Reducing user complaints by 50%
57% of the identified issues improve energy efficiency
70% of visual inspections replaced by data analytics
Creating an interconnected ecosystem benefitting everyone on site, which also fights climate change.

Analytics making maintenance easier

Replacing maintenance tasks using analytics and a data-driven model can really improve productivity and be the basis for much more cost-effective facility maintenance. And improved productivity isn’t the only benefit. Other benefits include considerable energy savings and improved end-user satisfaction. With analytics running 24/7, issues can be identified more quickly and before they impact end users. Weekly checks, which are commonplace, will not pick up these issues at the same rate. An analytics and data-based approach to facilities maintenance and asset performance services not only provides greater convenience for the end user, but also makes maintenance easier and ensures better use of resources.

Panu Kärävä is a global portfolio manager for energy and asset performance services at Siemens Smart Infrastructure.