Data Center Water Monitoring: Omen’s $16M Deal
Omen AI’s $16 million Aquatic Labs deal combines fluid sensors to help reduce cooling risk, protect data center uptime and manage AI’s water demands.
Why Omen AI Is Betting on Better Data Center Water Monitoring
AI infrastructure usually gets talked about in terms of chips, power, and capacity. Water doesn't make it into the headlines as often, but it's still part of the same equation. Cooling systems have to move heat away from increasingly dense computing equipment, and a problem in the plumbing can turn into expensive downtime fast.
U.S. data centers use an estimated 17 to 19 billion gallons of water a year for cooling, according to The Water Foundation. The group expects that figure to triple over the next five years. As more AI capacity comes online, operators have to manage cooling reliably - not just build cooling systems and hope for the best.
Omen AI is trying to solve one specific bottleneck in that process: the delay and uncertainty that comes with testing cooling fluids. Its $16 million cash-and-stock acquisition of Aquatic Labs brings together two sensor technologies that measure different properties of water. The plan is to combine them into a more complete, real-time view of fluid conditions.
This article looks at what the deal changes, how continuous monitoring could affect day-to-day operations, and what the acquisition can - and can't - tell us about the economics of AI infrastructure.
Why data center cooling water is becoming an AI infrastructure issue
AI makes cooling a business-critical concern. High-performance chips generate a lot of heat, and data centers have to remove it continuously while keeping equipment within operating limits. The cooling setup varies by facility, but in every design, fluid conditions and the system that moves it affect reliability.
Water demand is also getting harder to treat like a purely "facility-level" detail. The Water Foundation estimates that U.S. data centers use 17 to 19 billion gallons of water annually for cooling, and that use is expected to triple in five years. At that scale, operators, investors, and the communities around new facilities have practical questions: how much water is needed, how reliably it can be managed, and what happens to it after it's used.
For AI businesses, these questions matter because a data center isn't just a pile of servers. It's an operating system made up of power, cooling, network, and maintenance. If one part fails, compute capacity can be unavailable even when the chips themselves are still healthy. That's why cooling-fluid problems can quickly become an uptime and cost issue.
The operational cost of uncertainty
In the process described by Omen and Aquatic Labs, fluid testing often means sending samples to a lab and waiting several days for results. That delay slows how quickly operators can respond to changing conditions. And a test result reflects what was happening when the sample was taken, not necessarily what's happening right now.
The fallback can get expensive. If operators think the fluid system is out of specification, they may decide to dump and replace the water rather than risk damage to high-value equipment. That can bring cleanup costs and operational interruption while the system is restored.
Real-time measurement doesn't eliminate risk. Sensors still need to be accurate, maintained, and built into a workflow that staff can act on. But having faster information can help operators tell the difference between a problem that's still developing and a situation that calls for a full response.
Water management extends beyond the facility
There's also a broader water-management question. Aquatic Labs CEO Allan Adams argues that you have to measure what happens between water entering a data center, leaving it, and reaching municipal systems - and do it responsibly. Monitoring alone doesn't prove a facility is using less water or discharging it safely. It's a way to collect information that can support those decisions and help verify conditions.
That distinction matters. A sensor can improve visibility, but it's the operator's processes that determine whether visibility leads to lower use, safer handling, or fewer disruptions. For data center owners, the immediate payoff is operational. For communities and regulators, any larger value depends on what's measured, how the data is used, and whether it supports responsible water practices.
What Omen AI gains from Aquatic Labs
Omen AI and Aquatic Labs were both building real-time water sensors, but they approached fluid analysis from different angles. Omen's sensors focus on elements in the water. Aquatic Labs' technology measures alkalinity. Put together, the idea is that operators get a fuller picture than either measurement provides on its own.
That's the core logic behind the $16 million cash-and-stock acquisition: combine complementary sensing capabilities and teams, then integrate them into a single sensor. Adams described the benefit as a "much richer picture" of the water and fluid. The other piece of the plan is to replace a slow, manual testing workflow with more continuous measurement.
The deal also brings two different technical backgrounds together. Omen founder Zach Laberge previously built a company focused on fluid analysis for large construction machines that rely on hydraulics. Adams is a physicist whose work included advanced ocean sensing while he was an MIT professor and researcher at the Woods Hole Oceanographic Institution. He's joining the combined company as chief science officer.
Integration is the product challenge
Buying two sensor technologies doesn't automatically create one dependable product. The team still has to make the hardware work together, validate measurements in real operating environments, and present the results in a way data center teams can actually use.
A monitoring product needs to do more than answer "what does this sample contain?" It has to deliver reliable readings over time and help staff understand when a change is meaningful enough to act on. That can involve sensor calibration, installation choices, maintenance requirements, and integration with existing facility workflows. The announcement doesn't spell out how those details will be handled, so the real market impact will depend on what comes out of the integration.
Omen and Aquatic Labs expect the combined sensor to reach the market by early 2027. Until then, the acquisition is a bet on product integration and commercial execution - not proof that water use will drop.
What the deal's targets show - and what they do not
Omen told TechCrunch that its initial sensors would be monitoring chips pulling in 250 to 300 megawatts of power by the end of 2026. Laberge valued that installed chip capacity at roughly $5 billion to $7 billion. The company hopes to have sensors installed on the equivalent of one gigawatt of chips by the end of 2027.
Those numbers point to the scale of infrastructure tied to Omen's planned deployments. They don't, by themselves, measure water savings, customers served, or revenue booked. For operators evaluating the technology, what matters most is whether the sensors produce readings that drive action - and whether that action reduces avoidable water replacement, downtime, or maintenance costs.
How real-time water monitoring could change operations
The business case for monitoring is fairly direct: better information can help keep cooling systems within acceptable conditions longer, spot issues earlier, and avoid unnecessary shutdowns or fluid replacement. The value isn't just the reading. It's the decision it enables.
A simplified workflow looks like this:
- Measure fluid conditions continuously rather than relying only on samples sent away for testing.
- Detect changes that could indicate a developing issue, such as bacterial growth or a shift in alkalinity.
- Interpret the signal in context with operating conditions and maintenance procedures.
- Pick a proportionate response, from investigation and targeted maintenance to replacing fluid if necessary.
- Check the outcome to see whether the intervention resolved the problem and whether the system can keep running safely.
Omen's sensors were designed to detect elements and potential issues such as failing valves or bacterial growth that could restrict water flow. Aquatic Labs adds alkalinity measurement. Together, the claim is that operators get more context about fluid condition than a single measurement would provide.
The economics are about avoided disruption
Data center operators make decisions based on the cost of failure. If fluid conditions become uncertain, replacing water and pausing operations can seem safer than risking damage to equipment. But that conservative approach comes with its own costs: labor, cleanup, materials, and lost operating time.
Monitoring can change that calculus by reducing uncertainty. If a sensor helps staff identify a problem early and figure out whether it's localized or systemic, they may be able to intervene before a full shutdown becomes necessary. If conditions stay acceptable, they may be able to keep using the same fluid for longer.
Still, this is a value proposition, not a guaranteed result. Sensors won't magically prevent every failure, and they can't replace good system design and maintenance. The impact depends on measurement quality, how quickly teams respond, and whether the cost of installing and running the system is lower than the costs it helps avoid.
What operators should validate before deployment
A data center evaluating a system like this should get specific before assuming it's a water-saving play:
- Which fluid properties does the sensor measure, and what operational risks do those measurements help identify?
- How often does it require calibration or maintenance?
- How are alerts separated from normal variation, and who is responsible for responding?
- Can the sensor fit into the facility's existing cooling setup and operating procedures?
- What baseline will be used to judge changes in water replacement, downtime, and maintenance?
The baseline is particularly important. Without it, an operator may end up with more data - but not necessarily proof that the system uses less water or avoids costly interruptions. A controlled rollout can compare outcomes before and after installation while accounting for changes in facility load and operating conditions.
What the acquisition says about AI infrastructure markets
This deal points to a quieter part of the AI infrastructure market. A lot of public discussion focuses on chips, electricity, and building new data centers. But once those facilities exist, operators still need systems that keep them running. Monitoring products sit in that operational layer, where reliability and maintenance costs matter even when the mood around new construction shifts.
That's how Omen frames the bet on resilience. AI companies and investors have discussed as much as $4 trillion in AI infrastructure spending by 2030, even as doubts about timing and returns have weighed on hyperscaler stocks. Laberge said Omen is aimed at customers building projects today or operating facilities that are already financed - rather than depending only on projects that may be funded later.
There's a practical distinction here. A slowdown in new construction could affect the timing and size of future demand, but it doesn't remove the need to operate infrastructure that's already in place. Existing data centers still need cooling, maintenance, and uptime. That means operations-focused products may be less directly tied to construction forecasts, though they're not insulated from budget scrutiny or changing customer priorities.
A market thesis, not proof of demand
Omen's acquisition is also a bet that specialized sensing can be integrated into infrastructure products with a clear operational payoff. The company isn't just buying a customer list or a finished mass-market product - it's combining capabilities and then asking the engineering team to turn that into something reliable.
For investors and potential customers, the checkpoints are straightforward: does the integrated sensor launch when expected, does it produce consistent measurements in real facilities, and do deployments lead to measurable changes in downtime, fluid replacement, or maintenance effort? Installed megawatts can hint at the size of a deployment opportunity, but they don't answer those questions.
The broader lesson for AI infrastructure is pretty pragmatic. Efficiency gains don't have to come only from more powerful chips or new cooling architectures. Better measurement of existing systems can also improve how capacity is operated. Whether that becomes a real business depends on whether the data changes decisions enough to justify the cost.
Conclusion
Omen AI's acquisition of Aquatic Labs targets a specific operational problem that sits behind AI growth: data center cooling systems need reliable information about the fluids they depend on. By combining elemental analysis with alkalinity measurement, the company is aiming to give operators a more complete, timely view than delayed lab testing alone.
But the deal only matters if it executes. The sensor has to work in real facilities, fit into maintenance processes, and produce measurable benefits - like fewer avoidable fluid changes or less downtime. For data center operators, that's the standard: evaluate monitoring by the decisions it improves and the costs it helps prevent, not by how much data it generates.