SUBJECT to the progress of the Railways Bill currently before Parliament, next year should see the formation of Great British Railways (GBR), the new public body reuniting infrastructure management with passenger operations under a single “directing mind” for the national network. In the meantime, work to prepare for the full launch is being undertaken by Shadow GBR, which brings together the future components of GBR: infrastructure manager Network Rail, some rail functions of the Department for Transport (DfT), and DfT Operator (DfTO) which is taking over passenger operations as contracts previously awarded to the private sector come to an end.
The lengthy process of rail reform has already borne fruit, however, with the creation in early 2025 of GBRX as the new strategic innovation body for Britain’s rail sector. The secretary of state for transport last year set accelerating the adoption of technology that improves efficiency, enhances passenger experience, and delivers real economic value as one of the five strategic priorities for Shadow GBR. “We’ve been given a clear and important mandate to break down the barriers to innovation,” says GBRX managing director, Toufic Machnouk, with the objective of enabling the railway “to adopt the technologies it needs to thrive in the decades ahead.”
Machnouk is able to draw on his experience in his previous post as the director of the Industry Partnership for Digital Railway at Network Rail, heading up the East Coast Digital Programme (ECDP) to install ETCS Level 2 on the southern section of the East Coast Main Line (ECML), used by fleets from 28 different operators.
When considering the principal barriers to innovation that GBRX has been created to overcome, Machnouk identifies three. The first is fragmentation along commercial lines, a result of privatisation and intended to encourage competition between private operators. “GBR solves much of this but it is still a complex ecosystem with long-life assets,” Machnouk says. “There is also fragmentation in how we do things, with no overall strategy. GBRX will bring coherence on the big technological tickets with purposeful, clear action.”
Machnouk similarly highlights “designed-in barriers.” These include the Network Code, the set of contractual rules incorporated into every track access agreement between Network Rail and an operator, covering those areas where all parties are obliged to work together to the same standards and timescales, such as making changes to infrastructure or rolling stock. Designed-in barriers also include safety validation of change, product approval, procurement regulations and intellectual property regulations (IPR). “All these things are there for a good reason,” Machnouk says. But at the same time, “this creates a web of barriers that are all designed, frankly, to prevent change,” creating the impression that “the railway is inherently resistant to change.”
In contrast, the third barrier to innovation is not to be found in regulations or contracts, but is rather the lack of strategic expertise and capability in the field of innovative technology, making it hard to identify what will have the most impact and what is the best technical and commercial path for adoption at scale in the field. “There is no capability at present in frontier technologies and how to deploy them,” Machnouk says. A strategy is required to develop this capability, “otherwise it is random, it is a free-for-all,” he says.
System themes
Workforce and skills form one of the five “system themes” that will guide GBRX’s work to overcome barriers to innovation and accelerate the development of key technologies. The GBRX function plan published last year says that Britain’s railway is characterised by an ageing workforce and siloed expertise, combined with a growing capability gap in emerging technologies as the need to attract the next generation into rail grows more pressing. Machnouk highlights the major demographic challenge facing the railway, with about 5% of the workforce leaving every year but only 1.5% joining.
Improving passenger experience, network operations, network planning and asset management are the four other major themes guiding GBRX as it works across the system to accelerate adoption of “big tech trends” such as AI, data integration, and what Machnouk describes as “sensor technology in its broadest sense, including computer vision where there are many applications,” as well as simulation and robotics. Onboard equipment will be a major focus, such as installing sensors to enable infrastructure monitoring to be undertaken by trains in passenger service rather than by specialised inspection vehicles. The creation of GBR should to a certain extent remove the commercial barrier between infrastructure management and passenger operations, but rolling stock or around 50% of railway equipment will still be owned by private rolling stock leasing companies, “a reality that is not going anywhere,” as Machnouk observes. GBRX will also have to work with other private entities remaining outside the remit of GBR, such as freight and open-access passenger operators.
We need to develop datasets that reflect the real world. That's where the hard graft is.
Toufic Machnouk, GBRX managing director
Reflecting wider trends to move equipment from the lineside to the train, Machnouk sees a major opportunity to replace train detection and location by lineside equipment or satellite tracking with quantum technology, measuring atomic, magnetic, or gravitational changes with a very high degree of precision to determine location, which is now being deployed in the maritime and aviation sectors as well as for autonomous vehicles. GBRX has joined with Imperial College London and other partners to develop high-speed quantum positioning technology, but it remains at an early stage of development. “This isn’t going to come next week but we could accelerate its development for rail applications,” Machnouk says.
No tech trend is bigger at present than Artificial Intelligence (AI), which Machnouk describes as “the most consequential technology of our time.” With investment by technology companies running into the billions, commoditised AI will become more prevalent, he says, but its application on an industrial scale in a safety-critical environment such as rail requires both careful consideration and a broader strategy. “Do we just let the market do this?” Machnouk asks. “Actually, no. AI requires guardrails.” To inform its work, GBRX has formed an AI Advisory Council to provide expertise in areas such as safety, ethics, infrastructure, data, and human resources. Members of the council include University of Edinburgh professor Shannon Vallor, an expert in AI and data ethics, and Dr Neda Naghshbandi, technical lead for intelligent infrastructure at AtkinsRéalis.
GBRX is leading the production of Artificial Intelligence in Rail: The Industry Action Plan, due to be published imminently. Machnouk says that it will set out the reality of industry readiness and the prospects for early deployment, as well as addressing fundamental issues such as the new regulations or other regulatory safeguards that will be required. The plan will contain 100 actions and identify pathfinder projects. “The challenge is finding the pathway to adoption,” says Machnouk, something that he sees as GBRX’s main purpose when tackling the barriers to adopting new technology, and where “the path is made by walking.” He summarises the GBRX approach as: “do we need to do something, how do we do it, how do we build expertise?” At present, the innovation funding and procurement environment is currently incentivised towards pilots and trials, but has no clear delivery path for adoption and scaling due to the complex combination of technical, commercial, regulatory and safety barriers identified by GBRX.
Among the potential uses of AI in rail, Machnouk highlights improving the passenger experience by creating an application that enables the customer to find the best fare, purchase a ticket and receive updates on their journey, including disruption to services, using spoken commands in simple English and possibly other languages. “We want to deliver that,” Machnouk says. On the operational side, AI could also help to restore the timetabled service following disruption, mining data that will need to be aggregated and standardised and which is sometimes still kept in manual format such as control logs. “We need to develop datasets that reflect the real world,” Machnouk says. “That’s where the hard graft is.” The deployment of high-performance computing will enable unstructured data to be used to develop predictive models to improve asset management and operational decision-making.
Methodology
GBRX’s core strength is the network of strategic relationships it has formed with operators, asset owners and technical experts within the rail sector, as well as suppliers and SMEs. Insight from outside the rail sector will be provided by researchers and sibling industries, as well as advisors and entrepreneurs. GBRX’s interface with the market is the X Platform, connecting “problem owners” with suppliers, SMEs and researchers, intended to provide support through early development and safety assurance.
As Machnouk stresses, innovation is not “trying random things and seeing what happens,” as wider society might see it, but requires a methodology. Here, GBRX has been inspired by leading innovation companies including X, established as Google’s “moonshot factory,” and has adopted the X Playbook to organise its capabilities to drive innovation in the real world. Problems are examined from first principles, with the hardest issue resolved first. Potential solutions are co-piloted with the problem owner, deploying a diverse range of capabilities relevant to the task in hand and deliberately shifting perspective at regular stages to provide fresh insight.
Silicon Valley may seem an unlikely inspiration for an industry often perceived as resistant to change, and Machnouk is clear that the structure of Britain’s railway is the major barrier to innovation. “The biggest challenge will be in the system,” he says. But change must come, particularly as more automation is needed to help the railway deal with the demographic challenge and insufficient capability in the skills of the future. “I want to be an agent of change, and help the industry deliver change,” Machnouk says. “Staying where we are is a disaster.”
