How many people are in the building on Fridays? What was the average temperature on the second floor last month? And how much energy does an employee use on average per day? ISS can now ask a building questions like these directly via chat – and receive a reply within seconds. But how does it work?
.png?width=1620&height=2025&name=Design%20ohne%20Titel%20(17).png)
Data-driven facility management is becoming the norm. Companies want to understand and optimize their buildings - in terms of energy consumption, user experience, and cost efficiency. That’s why ISS has been relying for years on sensors installed in buildings to monitor them in real time.
The collected data can be used to create “dashboards” that visually represent key metrics. At a glance, you can see the air quality in the building, how many workstations are occupied, or how often the restrooms have been used since the last cleaning. “Creating such dashboards is time-consuming, and the downside is that they aren’t dynamic. That means they only answer a specific question. If you later realize that it would have been more important to interpret the data differently - that is, to ask a different question - you have to start all over again,” explains Roger Gygli, Head of Digital Transformation at ISS Switzerland.
And that’s where artificial intelligence (AI) comes into play. For the past year, Roger’s team has been working with the Swiss technology company Akenza to develop a platform that interprets data dynamically. Where is the training taking place? At ISS headquarters, the CUBE. “Here, we collect 150,000 data points per day using 600 sensors. We’ve fed these into the platform so we can then ask it questions that train it to interpret the data. We reviewed the platform’s answers and then provided feedback on whether they were correct or incorrect. That’s how you train an AI.”
The result: a chat platform for ISS employees. They can communicate with the buildings they manage by asking, for example, “How many people are in the building on average on Wednesdays?” to optimize work schedules and cleaning routes. The answer comes in a flash. Customers benefit as well: They can analyze energy consumption, adjust space requirements, and reduce costs.
What was developed at CUBE can be applied to any sensor-monitored building. “I’m not aware of any comparable product on the market yet, which is why we’re playing a pioneering role in the field of artificial intelligence with this application. There’s interest from customers; we’ve prepared a proposal for a major project and hope that our AI will be put to use there soon,” reports Roger.
Roger Gygli
Head of Digital Transformation, holds an MBA from ETH Zurich and has been working for ISS Switzerland for 8 years.
