Skip to Main Content
Let's Talk
Menu Close
Category: Tech Talks

The Future of Building Optimization Is AI-Driven

The Future of Building Optimization Is AI-Driven

Published
24 November 2025

How AI Building Optimization Cuts Energy Costs Through Predictive Maintenance

By Sam Voisin, Senior Machine Learning Software Engineer at Optimum Energy

Across every sector from healthcare and higher education to advanced manufacturing, leaders are rethinking how their buildings perform and the cost of that performance. Energy optimization has long been about reducing waste and improving efficiency, but today the stakes are higher. Achieving true performance resilience now depends on a critical capability: artificial intelligence. 

Beyond Control, The Path Toward Intelligent Optimization Economics

Traditional building optimization relies on set schedules, static thresholds, and reactive adjustments. AI changes that model completely. By continuously analyzing real-time data from HVAC systems, peripheral sensors, and external data sources like weather forecasts, AI transforms static operations into dynamic systems that learn, predict, and adapt. 

This continuous learning is what makes AI so powerful. Intelligent analytics tools evolve with every data point, understanding how climate drift and operational changes impact performance over time. They learn the critical factors affecting facility performance and adjust accordingly, ensuring systems operate efficiently even under shifting conditions. Through predictive analytics and sensitivity analysis, AI can evaluate how different efficiency measures respond to changes in the local climate, allowing for proactive adjustments that minimize costs and energy use. 

Predictive Maintenance and Continuous Reliability 

AI-driven systems do not wait for equipment to fail. They identify early warning signs and adjust operations or alert teams before issues escalate. This shift from reactive to predictive maintenance reduces downtime, extends equipment life, and preserves capital budgets while maintaining comfort and reliability.

“For as long as I’ve been in the industry, the holy grail of smart buildings and facility management has been trying to switch from reactive, break/fix management to predictive maintenance. AI is finally putting this within reach and Optimum is primed to capitalize, sitting on over 20 years of accumulated performance data.” -Stephen Kozlen, Sr. Product Manager in Development

AI’s ability to aggregate and interpret data across an entire facility allows it to detect long-term trends in performance and equipment health that might otherwise go unnoticed. By recognizing subtle patterns that indicate wear or deterioration, AI helps ensure that equipment lasts longer and continues to perform efficiently for years to come.

A Smarter Path to Sustainability and Comfort?

AI also bridges the gap between sustainability and occupant experience. By understanding patterns of use, environmental conditions, and cost drivers, AI fine-tunes temperature, lighting, and ventilation in real time. It balances comfort and energy efficiency, optimizing the best possible outcomes for both people and the planet.

Because AI solutions continuously learn from data across Optimum Energy’s global portfolio of facilities, every system benefits from collective intelligence. Insights gathered from each facility can inform performance improvements across all others, driving faster adaptation and smarter outcomes on a global scale.

The Cost of Standing Still 

Many organizations are still relying on traditional optimization methods that are rules-based, manual, and disconnected from the full data picture. The truth is optimization without AI is no longer enough. As systems grow more complex and operational expectations rise, buildings that lack intelligent automation risk higher costs, lower efficiency, and lost competitive ground. 

According to EnergyInnovation.org, wholesale electricity prices are expected to increase by 25 percent by 2030 and 74 percent by 2035, while consumer electricity rates are projected to rise between 9 and 18 percent by 2035. With such significant increases on the horizon, facilities that do not leverage AI to continuously monitor and adapt will face escalating operational expenses and reduced resilience. 

Driving toward Realized Efficiency and Savings

AI is not the future of building optimization; it is the standard. Integrating intelligent analytics, predictive controls, and data aggregation into your energy strategy ensures your systems perform at their best today while continuing to learn and improve tomorrow.

With scalable, seamlessly integrated solutions, Optimum Energy delivers optimization that adapts with your facility over time, providing measurable savings, resilience, and operational confidence in a rapidly changing energy landscape.


About the Author, Sam Voisin

Sam Voisin is a data scientist and machine learning engineer with more than ten years of combined experience in industry and academia. He holds a master’s degree in Statistical Science from Duke University. Sam has developed and deployed machine learning applications across diverse sectors, including defense, healthcare, economics, and energy optimization. His work centers on building scalable AI systems that deliver actionable insights and measurable results.

Related Insights

Transforming Infrastructure: From Expense to Revenue Stream

Transforming Infrastructure: From Expense to Revenue Stream

Published
17 November 2025

Can Infrastructure Be a Profit Center? How Optimum Energy Can Turn Central Plants into Cash Flow

The Deal Podcast | November 17th, 2025

With Ken Bernhard, VP of Project Development


Ken Bernhard, VP of Project Development and fourth-generation construction industry professional, discusses our innovative approach to infrastructure development through energy-as-a-service models. The conversation covers how Optimum Energy transforms expensive capital expenditures into revenue-generating assets for hospitals, universities, and large manufacturing facilities.


Read the Full Transcript Here

Joshua Wilson:
Good day, everybody. Welcome to The Deal Podcast powered by FA Mergers. I’m here with a friend—a new friend—but I’ve known these guys for a while. Mr. Ken, welcome to the show. Tell us who you are and what you do.

Ken Bernhard:
So, I’m Ken Bernhard. It’s great to meet you and to be with you today and spend some time. I’m a developer—an energy developer—but not in the sense of oil and gas.

Joshua Wilson:
Right, energy developer not in the sense of oil and gas. Now, we’re here in Lafayette, Louisiana. I do a lot of work in Texas, and if you’re in Texas and say you’re in energy, people expect oil and gas—or maybe something involving the Dallas Cowboys, which I know you have a few stories about. So energy, but not oil and gas. Explain what that is.

Ken Bernhard:
Yeah, I don’t know much about oil and gas. I have a lot of friends who grew up in it, but I know very little. We’re on the energy conservation side. We do something called energy as a service, where we take infrastructure that’s primarily an expense and turn it into a revenue source.

These large central plants that run hospitals, higher education campuses, big universities, smart manufacturing—Intel, Johnson & Johnson, car manufacturers—they all have massive central plants they’re forced to build. But those plants aren’t revenue sources for them.

Growing up in the mechanical construction business, we built those plants. We were the experts. And we found a clever way to finance them off balance sheet and shift the risk onto the concessionaire. We did that, sold that company, and that went well. But then we found a better way to do it. So now I’m back a second time, doing it the better, smarter way—focused on long-term relationships and utilizing AI, because that’s where we are now.

Joshua Wilson:
Interesting. Okay, walk us through one of those deals. This is The Deal Podcast, so let’s talk deals. Two things I want you to go over:

Turning an expense into revenue — where did that idea originate, and how did your family execute it?

The clever way you financed things off-balance-sheet.

Walk us through both.

Ken Bernhard:
Sure. To get there, I’ll give a little background.

I’m fourth-generation in a mechanical construction business started by my great-grandfather. My grandfather expanded it, and my dad expanded it further. My uncle, who wasn’t part of the mechanical business, built a construction company from three people to 30,000. They built and maintained nuclear facilities, chemical facilities, military bases—that sort of thing.

When he sold his company, he came back to the mechanical business and said, “We did something interesting in the past—we built a big co-generation plant for LSU and Louisiana Tech. We were looking at one for Texas A&M. How do we do that again? And why did you stop?”

The answer was: the price of natural gas skyrocketed after Katrina, so it was no longer advantageous for universities. But their utility bills remained extremely high—1, 2, 3, even 7 million square feet of campus.

We explained the business model. He understood it immediately. Then he asked, “What stops you from doing it now?” And we said, “Easy money.” He replied, “I’m starting a capital company—so we’ll eliminate that. What else stops you?”

We realized nothing else stopped us. People just didn’t understand how to do it. So we set out to finance energy plants differently—by becoming a third-party concessionaire.

Here’s what that looks like:

We go into a large healthcare system. We long-term lease their central plant. In exchange for that lease, we perform a large energy conservation project inside their facility—maybe $25 million worth. That refreshes their infrastructure at a time when money is tight, margins are razor thin, and the political environment is shifting constantly.

After that, we sell back the chilled water, heating water, and steam that the central plant produces as a commodity—just like natural gas or water.

Auditing firms will classify this as off balance sheet if certain tests are met—one is having the ability to sell that commodity to neighboring properties within about three miles. As long as those tests are met, the hospitals keep the debt off their balance sheet and reserve their borrowing for revenue-generating projects like surgery suites and ER expansions—not for a $100 million central plant that only creates cooling.

But to us, the central plant is revenue generating.

We own, operate, and maintain it for 20–25 years. That was the opportunity. And it worked—it was wildly successful.

Joshua Wilson:
Okay, before we get to the “different,” let’s talk numbers. What does a typical range look like?

Ken Bernhard:
It depends on the size of the facility. The valuation of a central plant isn’t replacement cost—it’s more like real estate, or really, the output of the plant. What can it produce? For how long? At what price?

One question I always get: “Why would you reduce energy use if you’re selling chilled water as a commodity? Wouldn’t you want them to use more?”

Our thesis: No. Every central plant is oversized by 30–40%. There’s stranded capacity. If you reduce the building’s thermal appetite through conservation, you create even more stranded capacity—which you can then sell to neighbors.

That’s margin.

Joshua Wilson:
I love it. So the pitch is: we take an expense off your books, refresh your infrastructure, you pay the same or less for chilled water, and we assume the risk. And the hardest part isn’t the engineering—it’s trust.

Ken Bernhard:
Exactly. When you innovate, you immediately meet resistance. “This isn’t how we’ve always done it.” But once trust is established, everything else follows.

Joshua Wilson::
Now let’s go to the new way—AI. You sold the first company. Congrats. Now you’re approaching this differently. What does that look like?

Ken Bernhard:
The past company was named Bernhard. We were vertically integrated—design, build, finance. But to achieve off-balance-sheet treatment, we had to take control of the asset, meaning we operated it for 25 years. That tail was necessary, but our focus back then was really design and construction.

Then we discovered a small software company in Seattle—Optimum Energy. Their software dramatically reduces the cost of operating a central plant. Operators can do it manually, but not at scale. Optimum originally thought they were a pure software company—sell a CD, load software, run algorithms. But every central plant is different. Their software needed the plant to be “prepared” first.

They ended up becoming a tool for large ESCOs. Johnson Controls white-labeled it as “Central Plant Optimization 30%” because on average it saved 30% in utilities. Huge impact—40% of a campus’s electricity goes to the central plant.

Johnson did 140+ projects using it.

When we saw Optimum, we realized something:
They weren’t a software company.
They were an energy-as-a-service company—they just didn’t know it yet.

Today, Optimum’s AI augments operators. It doesn’t replace them—it scales them. One operator could never run 100 plants manually. Software can. We leverage the savings to refresh infrastructure and manage long-term performance.

We’re focused on the 25-year relationship—operations and maintenance excellence. Do a good job for the first 25 years, and you’re hired for the next 25. That’s a 50-year plan.

Joshua Wilson::
Your family has been building businesses for four generations. Why is a multi-generational mindset so important?

Ken Bernhard:
We’re owned by a family capital company—Bernhard Capital Partners. At the end of the day, what we’re building is a utility company. Mid-sized cities used to have their own utilities. Over time, infrastructure matures and must be rebuilt. We’re in the business of rebuilding—smarter.

Air conditioning hasn’t even been around that long. Two generations ago most homes didn’t have it. Infrastructure needs to be refreshed.

Joshua Wilson:
With data centers, EVs, crypto mining—all these new power demands—how does your model fit into that?

Ken Bernhard:
Conservation is cheaper than building more generation. The demand for power has never been greater, and the grid can only scale so fast. In many regions—especially the Northeast—the grid is fragile.

So the responsible approach is: use technology to do more with less.

Just like cars went from carburetors and eight miles per gallon to 30-mpg trucks and electric vehicles, infrastructure must evolve. We’re just helping it evolve.

About Optimum Energy
Optimum Energy is an established global leader in holistic energy infrastructure optimization, providing comprehensive and integrated solutions to mission-critical facilities. Its expertise spans a full spectrum of services, from initial engineering and construction to ongoing maintenance, asset management, and Energy-as-a-Service (EaaS) offerings. By serving key sectors such as healthcare, higher education, and advanced manufacturing, Optimum Energy delivers measurable efficiency and enhanced resilience. The company leverages proprietary innovation and AI-driven technology to provide continued savings, reliability, and resilience, while its flexible financing models, including off-balance sheet structures, enable its clients to accelerate cost savings and meet performance goals with confidence. For more information, visit www.optimumenergyco.com