Tech Giants Adopt AI for Sustainable Infrastructure Growth

The technology sector’s infrastructure crisis could be reaching a tipping point, and digital solutions are emerging as the answer.

On 14 April 2026, Veolia unveiled its Data Center Resource 360 offering at London’s Outernet venue, presenting a technology-driven approach to managing the resource demands that are increasingly constraining hyperscale operations. The French environmental services firm, which operates water and waste infrastructure across five continents, is positioning AI and predictive analytics as critical tools for an industry facing regulatory bottlenecks and resource scarcity.

The digital platform aims to help data centre operators achieve three primary objectives: carbon neutrality, water positivity and circularity. According to Veolia’s estimates, the technology could enable operators to reduce water footprint by up to 75%, improve energy efficiency by up to 20% and achieve waste recycling rates of up to 95%.

At the core of Data Center Resource 360 sits Hubgrade, Veolia’s existing digital platform. The system uses AI and predictive analytics to monitor water consumption, energy performance and maintenance operations in real time, providing operators with the data intelligence needed to optimise resource management across their facilities.

Hubgrade’s AI capabilities transform resource monitoring

The Hubgrade platform integrates machine learning algorithms that analyse historical consumption patterns alongside real-time operational data to predict equipment failures before they occur, potentially reducing unplanned downtime by up to 30%.

By processing data from thousands of sensors across multiple facilities, Hubgrade can identify anomalies in water usage patterns that might indicate leaks or inefficiencies, often detecting issues hours or days before they would become visible through traditional monitoring methods.

The AI-driven system also optimises cooling tower operations by analysing weather forecasts, facility load predictions and energy pricing data to determine the most cost-effective cooling strategies. This dynamic optimisation can result in substantial energy savings whilst maintaining the precise temperature controls that data centre operations demand.

Veolia has already secured partnerships with some of the technology industry's most prominent players. The firm is currently working with Google, AWS, TSMC, Samsung, Intel and Micron across more than 100 facilities globally, embedding its digital monitoring systems into the operational infrastructure of the world's leading hyperscalers.

These partnerships involve deep integration of Veolia’s monitoring systems with the hyperscalers’ existing infrastructure management platforms, creating unified dashboards that allow operators to view resource consumption alongside computational workloads and capacity utilisation metrics.

Richard Kirkman, Veolia’s CEO for Northern Europe, framed the technology’s value proposition in economic terms at the event. “Around 50% of the cost of AI infrastructure comes from water and power consumption,” he said.

A more efficient approach to resources and energy could cut the cost of running data centres significantly, he added.

“Having innovation deliver that resource efficiency is critical,” Richard explained at the event.

The scale of the challenge facing the technology sector is substantial. By 2030, the combined water consumption of data centres and semiconductor manufacturing is expected to equal that of 46 million people, roughly equivalent to the combined populations of the New York, Los Angeles and Paris metropolitan areas.

Infrastructure bottlenecks threaten expansion

The infrastructure challenges extend beyond simple resource availability. Many existing data centre facilities were designed during an era when computational demands were substantially lower, and the physical infrastructure often lacks the capacity to handle modern AI workloads.

Power grid connections represent a particular constraint. Securing sufficient electrical capacity for new hyperscale facilities can take years in some jurisdictions, with grid connection queues extending beyond 2030 in certain regions. Water infrastructure faces similar limitations, with municipal water systems in many technology hubs already operating near capacity.

Global data centre capacity is projected to nearly triple by 2030, with facility numbers growing at around 11% per year through 2034. The semiconductor industry is experiencing similar expansion, with chip manufacturing expected to grow by 26% in 2026, driven by a global trend towards domestic manufacturing as countries seek to reduce reliance on traditional electronics powerhouses like China.

However, resource constraints are creating regulatory obstacles. In 2026, nearly half of all planned or under-construction data centre facilities face potential permitting delays, driven primarily by concerns over water consumption and energy use.

Estelle Brachlianoff, Veolia’s CEO, highlighted the geopolitical dimensions of water scarcity at the event. “Water has erupted as a centre of conflict in the Middle East, important as oil, if not more,” she said.

AWS deploys recycled water systems

The event featured Will Hewes from AWS, who participated in a panel discussion alongside Veolia's leadership team. Will outlined AWS’s sustainability commitments, noting that the company has more than 700 carbon-free energy projects worldwide and is actively working to eliminate the use of potable water for cooling across its data centre estate.

AWS’s technology deployment strategy involves implementing advanced water recycling systems that can treat and reuse water multiple times within the same facility. These closed-loop systems significantly reduce the facility’s freshwater intake whilst maintaining the cooling performance required for high-density computing environments.

“We’ll have, in the next couple of years, over 120 data centres around the world using recycled water,” he said at the event, adding that AWS has also announced more than 45 water replenishment projects to date.

The deployment of these systems requires sophisticated water treatment technology capable of removing the minerals and contaminants that accumulate through repeated cooling cycles. AWS’s partnership with Veolia provides access to this specialised treatment expertise, enabling the hyperscaler to implement recycling systems at scale across its global infrastructure.

However, technology companies cannot implement such initiatives without specialist expertise. “We’re really good at operating data centres but we are not water treatment or wastewater treatment operators,” Will explained at the event. “We have some really complementary skillsets here with Veolia that can help us expand our ability to use recycled, instead of potable, water.”

Richard emphasised the importance of deployment capability in technology partnerships at the event. “We’ve got the experience of not only developing new technologies but also that 'deployment magic'. That's often where projects fall down and dreams fall apart,” he said.

“Our mantra is to get the solution on the ground, running quicker, on time and affordably.”