Network Rail’s Data Backbone Gets a Seven-Year Technology Upgrade

Network Rail is placing a £133.2 million bet on the infrastructure behind Britain’s railway technology. The UK railway infrastructure operator has selected Cognizant to manage its data centre infrastructure, with services scheduled to begin in 2027 and continue for nearly seven years.
The contract reaches far beyond server rooms. It covers the networks, mainframes, cloud platforms, edge computing rooms, servers, physical data centres, and specified applications that support Network Rail’s operations. As AI and connected systems demand more power and stronger links, this infrastructure will sit at the centre of the railway operator’s technology plans.
A Major Contract With A Wide Technology Footprint
Network Rail selected Cognizant through a competitive flexible procedure under the UK’s Procurement Act 2023. Cognizant emerged as the sole successful supplier after all five bids reached the final assessment stage, beating Tata Consultancy Services, HCLTech, DXC Technology, and Atos IT Services UK.
The contract is valued at £133.2 million excluding value-added tax, or £159.84 million including VAT. The agreement is estimated to be signed on January 27, 2027, with services scheduled to run from January 29, 2027, to January 26, 2034.
A standstill period runs until November 23, giving unsuccessful bidders time to challenge the award decision. Once the engagement begins, Cognizant will manage the infrastructure and the tools and services needed to operate it, while Network Rail can also procure optional project services during the contract period.
- Physical data centres
- Local and wide area networks, including LAN and WAN
- Edge computing rooms
- Mainframes and servers
- Cloud platforms
- Specified applications
- Infrastructure operation tools and services
That list shows why the deal matters. Network Rail is not handing over one isolated system; it is bringing a broad layer of computing and connectivity under one long-term management arrangement.
Cognizant Brings A 13-Year Network Rail Relationship
Cognizant’s relationship with Network Rail spans 13 years and includes cloud migration, IT service management, and operational technology analytics. In 2020, Cognizant led a consortium with Amey Consulting and Arup to develop artificial intelligence, Internet of Things, and data analytics capabilities for railway asset management.
That five-year programme focused on monitoring more than 12,000 connected railway assets to identify possible maintenance issues. The new infrastructure contract places Cognizant in a position to build on that experience across the systems that connect data, computing, and railway operations.
The company has also been expanding its wider infrastructure capabilities. Cognizant completed its acquisition of Astreya, a provider of data centre and managed IT services, in June. Announced in April, the acquisition was intended to strengthen Cognizant’s AI infrastructure and managed services capabilities.
In July, Cognizant secured a healthcare engagement with US-based Centene that could be worth more than $500 million. TCS, one of the companies competing for Network Rail’s contract, secured a multi-year engagement to manage ABB’s global network operations in July.
Why Resilient Infrastructure Matters For AI And Edge
The pressure on data centre infrastructure is rising as AI systems move from experiments into business operations. Five years ago, a typical enterprise rack consumed between 5 and 15 kW. Today, AI racks commonly operate between 80 and 150 kW, while next-generation platforms are expected to approach 250 kW per rack.
That jump changes the demands placed on power, networks, cooling, servers, and operating teams. Network Rail’s contract includes the systems needed to manage this connected environment, at a time when organizations are already delaying technology projects because their infrastructure cannot provide enough resilience.
Nearly three in four organizations, or 72%, delayed, scaled back, or paused an AI or edge project in the past 12 months over concerns about network resilience. Research from Opengear, a Digi International company, found that organizations with embedded network resilience reported estimated average outage costs of $12.4 million over the past 12 months, compared with $18.8 million for organizations with less developed resilience capabilities.
That difference amounts to $6 million in lower estimated average outage costs. The survey, conducted in July and August 2026, covered 514 CIOs and CISOs across the US, UK, Germany, India, and Australia.
- Organizations with embedded resilience reported 21% deployment delays of four days or more, compared with 39% for reactive organizations.
- 49% of embedded organizations expressed confidence during routine infrastructure changes, compared with 28% of reactive organizations.
- During a cyber incident, 93% of embedded organizations were confident they could isolate affected systems while maintaining access to unaffected infrastructure, compared with 74% of reactive organizations.
- 84% agreed that network resilience is becoming more important as AI and edge infrastructure expands, compared with 69% of reactive organizations.
“The industry has traditionally treated resilience as a response to failure. This research shows it is also a source of performance,” said Patrick Quirk, President, Opengear.
Quirk added, “AI and edge are becoming more prominent in the business, and every organization has the same question: can I see it, reach it and manage it?”
Network Rail’s agreement with Cognizant puts that question into a long-term operational framework. When services start in January 2027, the focus will be on keeping the data centres, networks, mainframes, cloud platforms, and edge systems connected and ready for the next phase of railway technology.
Based on
- Network Rail Awards £133 Million Infrastructure Contract to Cognizant — analyticsindiamag.com
- Why data centre cooling is about more than just temperature | The Independent — independent.co.uk
- China Railway Engineering Group | Articles, Insights & Updates | Fortune — fortune.com
- 72% of Organizations Have Delayed, Scaled Back or Paused AI and Edge Projects Over Resilience Concerns | AFP — afp.com




