Smart Facilities, Smarter Energy

A single prompt to an AI chatbot can typically take 2-5 Wh (watt-hour), whereas a Google search takes about 0.3 to 0.5 Wh. This shows that AI takes nearly 10 times more energy than Google searches. Top tech companies are investing billions of dollars in building new data centers. AI implementations are expected to force up to a 160 percent spike in data center power demand. Therefore, new innovative techniques to cool these spaces and high networking infrastructure are being planned to support this latest development. This means that facility management is only expanding to meet the needs.

Global data center infrastructure is growing rapidly, with 8,000 facilities currently operating worldwide and the highest concentration in the United States. Their energy demands are growing at an alarming rate, projected to consume 16 percent of total U.S. power by 2030 — a dramatic increase from just 2.5 percent in 2022 before ChatGPT's launch. This surge highlights the massive energy footprint of an increasingly AI-driven digital infrastructure. With an increase in new data centers, the opportunities for FMs, energy consultants, network engineers, and electrical and mechanical engineers will be in demand.

It is a paradox: People plan to build by consuming energy and work to build an energy conservation plan. The development of AI is revolutionary. On average, AI takes six times less effort and time consumption. Because there is no turning back, the question is: how can AI help save energy?

As contradictory as it may sound, AI can be used to conserve energy. It begins with monitoring energy consumption. Tracking energy can provide an overview of the total energy and insights into what takes up the most energy. This data can be used to analyze how energy can be saved. Zachary Henry from NBC Universal believes that energy consumption data should be tracked at least monthly. It should be continuously tracked and data to be collected with no more than 15-minute intervals or change of state. He urges that to meet any sustainability goals; it is essential to establish a baseline and then measure the impact of changes on energy consumption. AI Agents or tools can be implemented to better run facilities. Here are some areas in which AI can be applied for maintaining and managing facilities:

1. AI in predictive maintenance

Unusual energy spikes can be determined during energy consumption tracking. Understanding the root cause behind the high energy utilization can help prevent equipment damage. Setting up an automated notification system to inform unusual energy patterns can help reduce the human effort behind the monitoring. AI can also be used to perform a root cause analysis with the energy, maintenance history and equipment information to predict the cause of the energy spike and resolve the issue.

A digital twin is a virtual replica of a physical asset, system or facility that uses real time data from sensors, IoT devices and software models to simulate, analyze and optimize performance. When integrated with artificial intelligence, it becomes a powerful tool for transforming facilities into smarter, more efficient and adaptive environments. Creating digital twins of large buildings helps in finding accurate square footage to calculate precise carbon footage. They are also useful for identifying underused areas and repurposing these spaces to enhance higher utilization. The use of live sensors can predict equipment failure. For instance, before a leak occurs, the sensors can alert the technicians and FM team. Similarly, cracks, corrosion and breakage can be predetermined and prevented through sensors and AI-powered cameras.

2. AI in analytics & decision-making

AI can assist in finding trends, relations and patterns with service requests and other resources utilization. Similar types of repeating service requests can indicate an issue or consume more energy than needed. For example, some hospitals have a lot of too-hot and too-cold service requests, indicating they need to maintain the optimum temperatures in all indoor spaces. This directly translates to energy waste. Tracking temperature and autonomously changing based on occupancy, space type, time of the year and day, and square footage area can be a game changer in the world of facility management and energy conservation. Monitoring service requests for similar entries and focusing on these specific areas can resolve issues easily. Analytics can further be used in preventive maintenance and corrective maintenance work order history to understand technician’s efficiency. Insights can be drawn from resource utilization, verifying if all technicians are busy and understanding whether the current staff is under or overutilized.

3. AI chat agent integration

Thinking about staff utilization, another area to save human effort is to implement an AI chatbot. A virtual chatbot integration with an existing computerized maintenance management system (CMMS) helps site managers and customers get their questions answered instantly. The end user does not need to spend time searching for information. An AI chat agent that can answer questions about critical assets and important work order history is useful for customers and managers. This will also help in getting information such as risk criticality, asset classification and work order type. The current AI chatbot models work well with unstructured data (email, free text) compared to structured data (table, Excel, Database).

AI can extract data and insights from unstructured data such as maintenance logs, reports and technician notes. Providing the maintenance history can make the AI perform a root cause analysis and give prescriptive maintenance solutions. It can quickly analyze patterns across multiple data sources. The following details can be provided to a chatbot to analyze maintenance predictions.

Field	Why It’s Useful
Asset ID / Equipment Name	For tracking failure patterns per machine
Work Order Date	Helps identify frequency/timing
Failure Description	Key for recognizing symptoms
Cause Description	Helps trace actual causes
Corrective Action Taken	Shows whether fixes are permanent or temporary
Technical Notes	Hidden insights often lie in free text
Work Type (PM, CM, EM)	Helps assess preventive vs. reactive work balance
Parts Replaced	Repetitive replacement points to deeper issues
Downtime Duration	Prioritization Metric
Cost	Economic impact of repeated failures

 

4. Automated work order management

AI can prioritize work orders based on urgency, asset condition and historical data. If a hospital’s HVAC system is showing early signs of failure, AI will automatically prioritize it over a noncritical lighting issue. AI chatbots can assist in retrieving manuals, standard operating procedures or previous repair records instantly. If information about the technicians is fed into the AI agent, it can automatically assign tasks to the right technician based on skill level, availability and workload.

Another key feature of AI is voice command. Technicians can use voice commands to create, update or close work orders hands-free on the go. Additionally, AI can track the status and progress of each work order in real time, sending automated notifications to relevant stakeholders, such as FMs or department heads, keeping everyone in the loop. AI can also generate detailed performance reports, providing insights into asset performance, technician efficiency and overall system health. With continuous learning from the data, AI systems improve over time, further optimizing work order management by identifying trends and suggesting preventive maintenance before issues arise. This leads to reduced downtime, more efficient resource allocation and long-term cost savings.

5. AI in regulatory compliance

AI agents are also useful for tracking regulatory compliance through automated reminders and flagging overdue preventive maintenance or inspections to maintain compliance. AI Agents can act as interactive training assistants, providing instant answers to compliance-related queries. They can help new employees quickly adapt to regulatory protocols in maintenance. AI can retrieve compliance records, generate reports and assist during audits. This minimizes the risk of missing critical deadlines, significantly reducing human error in tracking compliance activities.

AI can also continuously monitor regulatory changes, automatically updating compliance protocols and ensuring that the organization stays aligned with the latest standards. By automating compliance tracking and reporting, organizations can reduce the need for manual oversight, saving time and operational costs. The proactive nature of AI in anticipating and addressing compliance issues helps avoid costly penalties, and the streamlined process ensures regulatory requirements are met efficiently and consistently. With AI assistance, teams can focus on core responsibilities, while compliance remains a seamless, integrated part of daily operations.

These are some of the ways in which AI can save effort, time and money in FM. While using the publicly available AI chatbots is a great way to tackle daily tasks, they come with caveats. First, private and sensitive information cannot be added to these chatbots. The second issue is the limitation. Chatbots work based on rules and might not have specific knowledge about FM or the organization. Implementing AI and integrating it with the organization’s CMMS will be valuable and cater to their specific needs. It is crucial to conduct thorough research, identify the options that best meet the requirements and assess the success rate of implementation by benchmarking against top-tier organizations.

In conclusion, for FMs, there may be pressure to adopt AI out of fear of falling behind, but that should not be the case. AI is simply a tool designed to assist humans in performing tasks more efficiently. It is easy to get overwhelmed by the rapid pace of development, but there is no need to rush. Take the time to fully understand this technology and evaluate how it can enhance the current workflow.

AI should be integrated thoughtfully and strategically, with clear goals and an understanding of how it will align with existing processes. Rather than jumping on the bandwagon, take a measured approach to assess its true value. Continuous learning and experimentation are key to staying informed. Being adaptable allows one to make the best use of AI as it evolves.