You are here

IT Services, Contributed

How Machine Learning Enhances Facility Management

September 12, 2019

As the facility management (FM) industry grows at a lightning speed, it is becoming more complex and energy-wasting. Companies can focus on anomalies and make their processes more sustainable by leveraging machine learning (ML) as a new approach to data management. 

Machine learning (ML) technology powered by big data has applicable use cases within the facility management domain yet only a few businesses understand their value. 

Machine learning in facility management (FM) approaches can provide a powerful tool for decision making, fault prediction, and data management processes. Thus, ML technology powered by big data in facility management can automate a business’s routine tasks, increase the efficiency of services provided, and heavily reduce costs.

3 Ways Machine Learning Technology Can Enhance Facility Management

  1. Make smarter decisions driven by data 
  2. Monitor for system vulnerabilities and predict challenges 
  3. Strategically optimize data

How Companies Manage Spaces Today

Research and Markets reports that the value of global FM market reached $35 billion in 2018, and it is expected to grow to $59.3 billion by 2023.

Drawing from these statistics, FM is one of the fastest-growing industries in the world. 

Think for a minute what impact it exerts on energy and environment in general. Not that easy to grasp mentally, right? 

According to In-Building Tech, the average building wastes 30% of the energy it consumes due to built-in inefficiencies. 

Energy usage by commercial buildings comes with a price tag of $190 billion. Yet, 30% of this goes to waste costing $57 billion a year.


Apparently, without technology, the space management sector loses costs it could save and invest. 

To change this status quo and cut on energy bills, tech giants such as Google and Microsoft tackle the challenge of energy mismatching for their server buildings by using the ultimate ML advancements. 

ML is among the current trends in FM automation. It helps facility managers become more proactive when process failures come into play as ML tips the balance in their favor. 

Space managers are motivated to leverage machine learning in facility management and transforming data from IoT devices into actionable insights because they want to set new quality standards. These new standards will make it possible to:

  • Make continuous system health checks 
  • Improve operations 
  • Minimize energy waste

So how does it work? Penetrating the facility management system, machine learning is capable of providing computers with data insights powerful enough to make decisions automatically in the future. 

Typically beginning with a predictive algorithm, ML grows its potential exponentially as the computers adapt through experience. As a result, computers are learning to think on their own. 

Feeding the algorithms with more data, you can have intelligent machines make recommendations based on what they see. 

Ideally, a machine learning algorithm would automatically react to real-time conditions and provide a solution to the problem at stake based on historical data. 

Thus, business owners can address all factors that affect building operations such as forecasting challenges, detecting anomalies, and even normalizing operations against the effects of weather. 

3 Ways Machine Learning Benefits Facility Management Systems 

As business owners strive to optimize processes, prepare facilities to meet the demands of a more sustainable future, and spur intelligent decision-making in their spaces, they have started to leverage machine learning in different industries. 

At the end of the day, ML improves quality and performance standards of facility managers without little downtime. Additionally, machine learning helps them make smarter data-driven decisions, detect faults, and enhances smart data optimization in facility management systems. 

1. Make Smarter Data-Driven Decisions

What machine learning teaches computers is to brush up on the decision-making process. When combined with using predictive analytics, machine learning can be implemented to create building trends and patterns for different data streams and KPIs, which results in the optimization of resource utilization

Also, these technologies provide a possibility to generate new value from existing data, enabling facility managers to rely on them when making their decisions.

Data-driven decision making may follow a certain trajectory:

  1. External data (e.g., outside air temperature) is collected by an AI algorithm.
  2. AI with ML technology guides heating, ventilation, and air conditioning (HVAC) control decisions.
  3.  AI connects to the building management system (BMS) through an Internet of Things (IoT) gateway.
  4. BMS data and controls are fully leveraged. 
  5. HVAC setpoints are dynamically optimized for pressure and temperature. 

External data is gathered and then used by AI to guide HVAC control decisions.

The result is an optimal comfort level for every zone of the building with minimal waste. 

What decisions exactly can machine learning improve in facility management? 

There are many questions machine learning can answer such as: 

  • Which machines should be repaired or replaced to achieve better results?
  • How can expenses be allocated to achieve the greatest return on investment (ROI)?

In fact, machine learning has the capacity to address the information flow a human brain can hardly encompass.

2. Monitor Faults and Predict Challenges

Another benefit lies in fault monitoring. Machine learning allows detecting when facilities are operating at a non-optimal level and preventing expensive downtime that can have a detrimental effect on any business. 

Such an approach helps facility managers save time and money by conducting predictive maintenance of their spaces and assets (e.g., by ordering required components or replacing their equipment right from the very first sign of its weakness, before it actually fails).   

Transformations Undergone for Fault Detection and Diagnostics: Measurement, Features, Decisions, and Classes


When you use machine learning in your facility management system, you’ll be able to spot all kinds of anomalies, as it is not limited to predetermined markers. 

Based on a recorded historical set of data, ML identifies any aberrations in the system, considered not normal, and notifies operators. 

It can be a great addition to managing heating, ventilation, and air conditioning (HVAC) to create environment comfort in a way that was not possible before. 

3. Strategically Optimize Data    

Machine learning approaches can transform data into valuable insights. ML algorithms can automatically analyze, sort, classify and categorize data based on its type and the processing approach required. 

This enables facility managers faster to find the information they need and improves facility management in general.

In every building, there’s a blizzard of operations managers have to deal with, from HVAC to lighting. The benefit of machine learning is that it can establish better coordination of building systems that are becoming more complex with the course of time. 

With machine learning, operators can better understand how various systems interact.

Case Study: Using Machine Learning in Meeting Rooms 

How facility management benefits from machine learning can be shown in one of the real-life cases. Using predictive analytics for facility management at office rooms to maintain an optimal temperature, you can test the edge of ML-powered smart air conditioning system. 

Depending on the number of meetings conducted, the system is supposed to handle the air cooling efficiently.  

Because offices typically have a number of conference and meeting rooms, these spaces require air cooling during the day. 

Apparently, it is inefficient to constantly cool a meeting room due to the energy loss, so the system should stick to a certain schedule and activate when the meeting is due to start. This is where machine learning comes in, offering the following approaches: 

Predict How Many Events Will Take Place and the Number of Attendees

Information from the event calendar is not always enough for this task because meetings are often not scheduled in advance.

The forecasting is based on the historical data of meetings conducted in a particular room. 

In turn, visual person detection algorithms are applied to detect unscheduled events. For example, if more than two people are in the room, then it is considered to be a meeting.

Set Optimal Air-Cooling Mode

This is done with respect to the weather condition and expected meeting parameters, such as the time and the number of attendances, as well as other crucial factors. 

The approach powered by machine learning allows maintaining the optimal temperature in meeting rooms and reduce power consumption for air conditioners. It makes for a good example of how spaces can and should use ML.

Machine Learning Can Help the Facility Management Industry Guarantee Estate Efficiency

The role of facility management is to enhance estate efficiency, guarantee operational continuity, and merge legacy buildings and their systems as new facilities amplify.

To make buildings more intelligent and less energy-wasting, facility managers can leverage IoT and machine learning.

The application of smart data management solutions will ensure that builders, manufacturers, and operators save time and money, respond fast to critical anomalies, and optimize data to make predictions for the future.