Insight 27 November 2024

Accurate data: the foundation of good AI for facilities management

AI can be an invaluable tool to help facilities managers get the most from their estates. But to be truly useful, it needs to be fed with accurate and meaningful data.

By Dan Blake

Digital transformation is permanently changing the way we manage estates, and now AI is taking that to the next level. We’ve already outlined the potential benefits AI can bring to facilities management. Now, we’re discussing the most important factor in making sure it can deliver those benefits: data.

This blog will talk about the steps we took to build a solid data foundation for AI, share best practices from our own experiences and explain the value our approach to data brings to our clients.

An investment in the future

If you want to unlock the transformative potential of AI for your estate, you need a solid foundation of data – one which is robust, accurate, reliable and appropriate for your intended use cases.

At Mitie, we’ve invested seven years and over £160 million in our data infrastructure. Because we self-deliver across all our services, we have one of the market’s most in-depth structured and unstructured data sets, providing insight into engineering callouts, security incidents, time-to-clean, energy consumption and more. And because it’s data we generated, we know its strengths and weakness and in turn how we can best utilise it. Over 700TB of it.

But having all this data wouldn’t be useful unless we had the technologies and frameworks to ensure we could put it to use. Ensuring this took a great deal of strategy, investment and time. But it’s now the cornerstone of everything we do at Mitie.

The cornerstone of artificial intelligence

Today, all our data is stored in our state-of-the art data lake. We think of it as the “distribution centre” of our AI activities. Data from a wide range of sources comes into our central ecosystem, is structured, modelled and then efficiently distributed for reporting, analytics, automation and AI.

When we started building the data lake seven years ago, our goal was to establish large, reportable datasets that would give us a real-time understanding of what is happening within our business and client environments – and provide visibility over a range of critical domains.

Crucially, we also needed to make sure we had the processing power to transform our and structure our data at scale. We needed to be able to handle huge volumes of transactional data to provide the services that would make us a leader in facilities management AI.

To help with this, we’ve partnered with industry leaders in various fields, taking advantage of solutions like Microsoft Azure and our Cloudera Data Platform, as well as incorporating tools like Microsoft Fabric to provide different modelling and AI capabilities to our data analysts.

It cannot be overstated how important all of this is to our use of AI. Without this investment all those years ago, we wouldn’t have a foundation that enables us to explore and experiment with use cases in a meaningful way. It’s a first step that simply can’t be overlooked, because like all data-driven activities, the quality of AI hinges on the quality of the data you put in.

This is the “how” behind our AI activities. But what do we actually do with that data?

Summarise, recommend, predict and automate

By combining this dataset with AI, we are unveiling new opportunities to put data to work. The technology’s ability to process large amounts of data at speed provides rapid insights to our colleagues, enabling them to make more informed decisions for themselves, our company and our clients.

We use AI to summarise data, recommend best actions, predict events, detect anomalies, diagnose issues and provide more effective solutions – including automation.

When it comes to how you want to apply AI, the sky really is the limit. But here are just some of the ways we’re using it alongside our dataset.

To gain visibility

Gaining visibility into the way an estate operates is the vital first step in facilities transformation. To make meaningful changes you need a strong understanding of where risks, waste and opportunities lie.

Using AI and advanced analytics techniques, we’re able to combine and process huge amounts of data at scale, producing a complete picture for our clients that includes insights into assets, carbon hotspots, security, cleaning requirements and much more.

With these approaches, we’re able to dive deeper into these insights and spot patterns the human eye can’t detect. But this isn’t just about technology. The results are always fed to our human experts so we can find the perfect balance between state-of-the-art technology and the extensive experience of our people on the ground.

Reduce costs

One of the benefits of these insights is how easy it makes it to identify opportunities to remove cost. With data about how buildings are used, you can automate things like heating and lighting based on real-time activities, turning things off when they’re not in use to reduce costs and emissions. You can also deploy demand-driven services, like cleaning and engineering – maintaining exceptional standards without ever overprovisioning resources.

For many, deploying AI is a balance between investment and reward. You have to invest in the infrastructure to get any return. But as we’ve already made that investment on our clients’ behalf, we can provide the benefits without them having to foot the expense or tackle the complexity.

Create exceptional spaces

In the modern office, convenience is king – it’s what will get people back into the workplace and keep your building users happy and productive.

AI-driven applications can be used to simplify everyday office activities – from room bookings to concierge services. Applications with features like chatbots can simplify processes, both for those that use your buildings and those that manage them. And real-time visibility into the use of space – and usage patterns over time – can help simplify scheduling and define strategies for all activities.

Ensure security

Our state-of-the-art intelligence services process huge amounts of information and predict potential threats to your environments. Our Merlin Protect 24/7 business intelligence software harnesses millions of lines of data from multiple inputs to provide real-time insights into your threat landscape – so you can optimise the processes that keep your people, places and information safe.

Some lessons we’ve learned along the way

In building the foundation for these AI-driven capabilities, we’ve learned a lot about robust data management and providing a data resource you can trust. Here are some best practices to follow for your own data journey:

  • Make sure your data systems are well designed, aligned to governance standards and appropriately managed. This should encompass robust design prioritising master data, continuous improvement processes addressing user issues swiftly, and vigilant monitoring of interface issues.
  • Provide the training and education that guarantees everyone interacting with your systems understands how to use them to get the most accurate results.
  • Monitor compliance across your entire data ecosystem – and stay abreast of changes to legislation. Ensuring data is used appropriately and ethically is key to winning confidence with stakeholders and customers alike.
  • Regularly update systems to make sure data is continuously captured across your entire estate.
  • For more advanced AI and machine learning models, be aware of the risk of drift and bias. Be transparent about these risks and continually train your models to mitigate them.

Having the correct governance structures in place across your organisation is crucial for good data management. At Mitie we have a Data Governance Council, an AI Use Case Board and an AI Ethics Council that sit alongside technology and systems forums. These ensure the correct controls are in place to properly manage risk, remove bias and follow the leading best practices for data management and AI use.

If you’d like to learn more about how you can leverage our data lake and AI capabilities to optimise your facility, get in touch. Or read more about how Mitie is redefining the future of FM with AI.

Redefining the future of FM with AI

We believe that AI makes things more humanly possible. But what does this mean when it comes to delivering exceptional services? Discover the five ways we’re using artificial intelligence to drive true facilities transformation in our practical manifesto for facilities AI.

Find out more

Read next

See Resource Archive