The artificial intelligence (AI) that has been working quietly behind the scenes for years has surged to the forefront as businesses across industries try to figure out how to leverage AI to work better, faster and smarter.
AI is not actually all that new. The origins of machine learning and artificial intelligence date back to the 1950s. What is new is the introduction of Generative Pre-trained Transformers (yes, the GPT of ChatGPT), which are effectively powerful large language models.
“It’s only recently that this new generation of generative AI has come to the forefront and shined a spotlight on AI as a potential revolutionary new toolset,” says David Franklin, Industry Principal for Commercial Real Estate and AI at Yardi.
Commercial real estate owners and managers are still in the early stages of adopting AI into building operations and management processes. According to the 2024 AppFolio Property Manager Benchmark Report, 21% of more than 5 000 property management professionals surveyed said they currently use AI — and an additional 28% plan to adopt it. The most common application for AI is tenant communication. However, there is a proliferation of practical use cases for property owners and managers in core areas that include:
- Energy efficiency
- Building automation systems
- Tenant experience
- Security
- Occupancy tracking
The Building Owners and Managers Association (BOMA) has written a guide to inform property managers on the growing business case for AI, as well as practical steps to help develop AI strategies, smooth implementation, and the role that humans play in keeping AI on track.
Evaluating return on investment (ROI) and the business case
People like to talk about the cool things that AI can do, such as its ability to write emails and answer complex questions. But for property managers, the focus is on developing a business case to show how they can leverage AI to add value.
“Particularly in property management and building technology, it’s all about that ROI that you can achieve with it,” says Benjamin Crowe, Global Head – Product and Portfolio Management, Digital Service Portfolio at Siemens Smart Infrastructure. “We now see AI becoming more productive in helping to boost all the same areas where property managers can save money in the past, I think that’s where you will see the uptake get higher and higher,” he says.
Siemens typically sees minimum savings of 10 to 15% associated with a broad range of AI applications depending on the use case.
Some of the key benefits of AI that can improve occupier experience and impact the bottom line include:
- Reducing the downtime of critical systems.
- Improving maintenance planning and extending the life of equipment.
- Improved safety and security.
- Reducing energy and water consumption.
- Improving productivity of operational staff.
- Reducing time and resources spent on troubleshooting problems.
- Prioritizing maintenance and service calls.
- Better experience for building occupants and employees.
An example: TK Elevator uses actionable AI in MAX, the firm’s real-time, cloud-based, predictive maintenance IoT solution that improves building management and maximizes elevator uptime. MAX speeds up troubleshooting and also allows technicians to be proactive in spotting problems before they occur. The company has seen up to 80% earlier failure detection and up to a 50% improvement downtime with its MAX solution.
“If we can detect that an elevator goes out of service before a human recognizes it, and if we also have intelligence as to what caused the problem and what we need to do to fix it, we have a big head start that reduces downtime,” says Jon Clarine, Head of Digital Services at TK Elevator North America.
Consider your options
The market is becoming increasingly crowded with bright and shiny new technologies that all boast some element of AI. An important first step – focus on the problem you are trying to solve or a specific area where you are trying to create more efficiency and then look at solutions.
“If you’re in property management, you have to ask yourself, ‘What am I trying to accomplish, and then what is the correct technology to use?’” advises Franklin.
For example, a large language model, such as a GPT chatbot or virtual assistant, is going to do a very different set of tasks than a machine learning algorithm that is trying to predict changes in rents or identify data fields to abstract a lease.
- Vendor Partnerships: An easy way to access AI technology is to work with providers that are already incorporating AI and machine learning into their solutions. In fact, you may already be using AI and not even know it. Technology that used to be at work behind the scenes is now moving more to the forefront with user-friendly interfaces, such as dashboards and mobile apps.
- Off-the-shelf options: There are off-the-shelf, prepackaged AI solutions that allow managers to dip a toe in the water without having to build their own custom solutions. And because of the arms race and rapid evolution of evolution, there is going to be a lot of interesting off-the-shelf options ahead that can create greater operating efficiencies.
- Custom solutions: In the decision of whether to build vs buy AI technologies, building your own systems can be more challenging and costly. The advantage is that it can create a competitive edge or differentiation in what peers are doing. One of the most challenging pieces is getting enough good data that is structured in a way that your AI model can utilize it.
“In a space that’s moving so quickly, partnering is what we’ve chosen to do because it’s the fastest path to the outcomes that we’re seeking,” says Clarine. “And especially in a fast-moving space, my recommendation would be don’t jump into bed with one supplier. Keep your options open.” When considering an AI partnership, it is important to address data ownership.
- When applying domain knowledge and placing it into artificial intelligence through a partnership, what happens to the insights that come out of that from artificial intelligence? Who owns those insights?
- Take steps to make sure that those insights aren’t passed on to a
competitor. - If a relationship dissolves over time for whatever reason, how can
you carry those insights on to another provider?
Start small and scale-up
Managers can – and should – dip a toe in the water first before jumping into the deep end. “When you talk about where to start, it’s important to start small, become more of an expert in implementing AI with the right partners for your business, and then grow that over time as you gain more and more confidence and transparency into what you’re doing,” advises Crowe.
One helpful guide is the 2 x 2 approach. Take your top two use cases that you think will involve the least effort and be the least disruptive to implement, and focus on doing those quickly to learn if they work or not. If it fails, the idea is that it can fail quickly and have a small impact. In parallel, gather research and input from stakeholders during the beta that can be used to transform the entire portfolio or the way the business operates.
Setting projects up for success
Broadly speaking, roughly 85 to 90 percent of technology projects fail or face shortcomings due to an overreliance on the technology itself to solve the problems and bring the right stakeholders together, according to Siemens.
“If you over-rely or have too high of expectations on what the technology can do without a proper plan and service for how to implement it, this is a very common failure point that we see with customers and that we try to support,” says Crowe. “AI can be very effective as a tool, but it can also amplify existing problems or even cause more confusion or sow distrust when it’s not used correctly.”
Advice from the experts
- Consider the logistics of implementing new AI technology with your existing IT architecture.
- Be aware of new cybersecurity risks.
- Develop a change management plan.
- Know the scope and scale of what you’re getting yourself into before trying it.
- Leverage your relationships to see what your peers are doing.
- In the fast moving world of AI, choose an approach that is sustainable and work with technology partners and providers that are future ready.
When it comes to recommendations for property managers who want to leverage generative AI, a key starting point is managing their data. AI applications are data driven. So, it’s important to have good “clean” and accurate information that is stored in a repository that is accessible. Managers have to start thinking more about the context of information. about the context of the information. How is that data being collected, verified, stored and maintained?
A common fear is that AI will replace human jobs. With new technologies such as the car, computers, the internet and smartphones – history shows that we end up with more jobs as a result of unlocking new capabilities. Humans, and the value of human direction, can be elevated when we use technology to eliminate manual tasks and have our experts focus on higher level tasks and cost-saving opportunities.
