Introduction to DfMA

Artist's impression of a Flow bridge concept to be manufactured offsite
Offsite manufacture is becoming increasingly popular in the building sector, and there is now a push to fabricate civil engineering structures in factory conditions too. The benefits are well accepted, according to Sam Stacey, challenge director for Transforming Construction at UK Research and Innovation.
“If you make things in a factory environment, you can do it much more efficiently, control the quality more effectively, reduce waste and improve safety,” he says. “It is natural to assume that it is more difficult to standardise and go offsite with transport assets, but our experience has been that it can be done offsite; it’s just a question of thinking about it differently.”
As Stacey points out, many bridges are already built using factory-manufactured components, such as precast concrete segments that can be stitched together onsite. However, even more savings can be made by adopting Design for Manufacture and Assembly – a process that is traditionally used in other sectors such as the car industry and consumer products.
This approach focuses on ease of manufacture and efficiency of assembly, and is being explored by Network Rail in the development of two new standard pedestrian bridge designs – ‘AVA’ for station environments and ‘Flow’ to replace level crossings. Both involve the use of non- traditional construction materials and are built using factory-made components that require minimum work onsite to install.
AVA

The AVA bridge has been designed as a system that can be configured to any site
The aim of the AVA project was to develop a standard pedestrian bridge design that could be used at any Network Rail station. Existing standard bridge designs have been in use for 60 years and do not meet modern accessibility requirements or enhance the station environment.
They also take far too long to construct and are expensive: it takes between nine months and a year onsite to build one, and less than 17% of the £3m-£4m project cost for each bridge goes on superstructure materials. The rest goes on expenses such as the planning and execution of railway possessions, crane hire, bridge transport, getting planning permission and site set-up.
The AVA project team – comprising Network Rail, Expedition Engineering, Walker Construction, X-Treme Systems, MTC, Hawkins/Brown, SCX Special Projects, Norman Foster Foundation and Atelier Ten – started working on the bridge in March 2020 and produced its first options four months later.
Anastasios Papapanagistou, a Network Rail senior engineer who was part of that team, says: “This was as early as Early Contractor Involvement gets. The contractor was involved from day one and we genuinely started with a blank sheet of paper with the aim of creating a step change in how you design bridges.”
Manufacturer X-Treme Systems was involved in the project from the start. It had never previously worked on a bridge but was experienced in making components for the aerospace industry using stainless steel. It introduced the other team members to its manufacturing environment and showed them the opportunities afforded by automation and digital technology.

A prototype AVA section consisting of two full-size 1.2m modules
The AVA bridge deck is a truss that can be assembled in 1.2m-long modules. The components that form the modules are laser cut from 3m by 1.5m stainless steel sheets, then folded and bolted together, with tolerances as low as 0.2mm. The truss modules can be configured to suit a specific site and fitted with cladding, a canopy, lighting and other mechanical and electrical (M&E) services before being erected as close to finished as possible. The team’s objective was that the entire structure could be installed onsite within a single 52-hour track possession.
During design development, X-Treme Systems built a section consisting of two full-size modules as a prototype. “That helped us to optimise how the stainless steel sheet was cut and folded,” Papapanagistou explains. The prototype also helped the team to check how easy the sections were to connect, and confirm that the proposed floor surface would not get clogged with dirt.
Eva MacNamara, associate director at Expedition Engineering, says the experience of having all of the team members involved from the outset was “the way you want projects to happen but don’t usually”. The manufacturer helped the others to understand what was within their capabilities, while having the client (Network Rail) as part of the team meant ideas could be tested against operational requirements and standards quickly.
AVA bridge concept
The aim of the AVA project was to develop a standard pedestrian bridge design that could be used at any Network Rail station. The bridge is designed as a system that can be configured to any site.
A full-size version of the AVA bridge will be built at Network Rail’s test facility in Nottinghamshire later this year, complete with lifts, which have been the subject of a separate DfMA innovation project. That will help to inform any changes that could improve the design, manufacture or assembly ahead of AVA becoming a standard design for use anywhere on the network.
AVA was part-funded by Innovate UK through the Transport Infrastructure Efficiency Strategy (TIES) Living Lab programme, which was set up by the Department for Transport to increase value and reduce inefficiencies. TIES Living Lab comprises 25 organisations, including High Speed 2, Network Rail, Transport for London and National Highways, as well as designers, suppliers and contractors. It focuses on DfMA and digital- and data-driven technologies, with the aim of cutting delivery times, boosting productivity, reducing carbon impact and creating safer sites.
Flow

Knight Architects was challenged to find a way of delivering a visually appealing footbridge that was both cost-effective and quick to produce
The Flow project addresses two issues faced by Network Rail: the need to improve safety at level crossings, and the desire for an alternative to the standard heavy-steel footbridges commonly used on the railway.
Network Rail wants to close and upgrade all 6,000 level crossings across the network, of which about 2,250 are footpaths across the railway that rely on pedestrians having to stop, look and listen for trains. The safest solution is to replace these with footbridges, but Network Rail currently has only one option: a standard design that is heavy, unattractive and expensive.
The organisation challenged Knight Architects to find a way of delivering a visually appealing footbridge that was not only cheaper than the existing standard design but also quicker to produce and suitable for different locations.
Network Rail wants to close and upgrade all 6,000 level crossings across the network, of which about 2,250 are footpaths
The result, Flow, is a modular system made from glass reinforced plastic (GRP) composite materials that can be configured in different ways to meet the requirements of individual sites. Flow was developed using a DfMA approach, with the manufacturers – KS Composites and Sui Generis – embedded in the team from the outset.
Also in that team were Network Rail, acting as both client and structural designer, independent (Category 3) checker Jacobs and key suppliers for the parapets, foundations and structural health monitoring.
Tom Osborne, international director at Knight Architects, says the team was determined to show that standardised design could meet standards of excellence and wanted to focus on the needs of the people who would be using the bridges. “When you are designing an object, the conversation quickly goes to the ‘what’ and ‘how’,” he explains. “We wanted to turn that on its head and focus on the ‘who’ and ‘why’, and let the design flow out of that.”
This is evident in the bridge’s design, which has curved bends between the stairs and the span rather than blind corners, and glazed parapets that flare outwards to improve visibility and provide a view over the tracks.
Flow project
Flow is a modular system made from glass reinforced plastic (GRP) composite materials that can be configured in different ways to meet the requirements of individual sites.
The bridge consists of a support structure – the ‘spine’ – and a separate curved deck comprising the main span and the stair units (or ramps in a fully accessible version), all of which are made from moulded GRP. The spine and deck are disconnected, so the curved deck can ‘flow’ around the corner while the supporting spine is orthogonally aligned to the railway, minimising the footprint. The spine acts as two arch-like structures.
The spine is installed first, then the deck modules can be lifted in incrementally in small, manageable sections that can be taken out and replaced if necessary. The spine provides rigidity between deck modules and can be connected very precisely to the concrete-free RapidRoot foundation system. The deck sections are fixed to the spine using a ‘gravity clamp’ system, avoiding fixings on the rail side of the bridge.
Bridge section
The deck sections are fixed to the spine using a ‘gravity clamp’ system, avoiding fixings on the rail side of the bridge

Service cables are integrated into the spine, so there are no cable conduits on the outside of the bridge, and structural health monitoring can be embedded to provide real-time data about the structure’s health.
A prototype of the Flow bridge was built in 2021 at the Long Marston Rail Innovation Centre in Warwickshire, where it was analysed and tested. That structure has now been taken down and is set to be installed at a crossing site in Wales.
Osborne says the approach adopted for Flow has resulted in many benefits in addition to the bridge being more attractive and user-friendly than the old design. These include it being lightweight, cost-effective, easy to maintain, quicker to install and more creative in shape. Other benefits of DfMA include bringing in supply-chain skills early, overcoming hurdles at the design stage, a greater focus on the user, a flexible design that can be reconfigured and the ability to continuously improve through asset monitoring.
The goal is to make the Flow bridge more sustainable, Osborne says – for example, by using more recycled plastic, natural fibres and recycled or natural resins in the GRP. The fibre optic monitoring in the structure will also help engineers and designers to understand whether there are places where the structure could be thinner, reducing the overall volume of material.
Key insights
Lessons learnt from the project include:
- The manufacturing process means each component is almost too perfect, which can present a challenge when it comes to fixing the structure to the foundations when you get to site. “You have to think carefully about tolerances and have a good understanding of the site,” Osborne says.
- The way the team is set up is crucial to the project’s success. “It is vital to make sure everyone is on the same page and the goals are set early,” he says. “That’s why the involvement of the client is so important – it stops you guessing and keeps the focus.”
- The process benefits from a non-siloed approach. “Breaking down traditional hurdles – for example, doing the design and then passing it to construction – allows the process to be more iterative,” Osborne says.
- The industry’s mindset about bridge design needs to change. “The default way of thinking is that you either have an entirely bespoke solution or you have an off-the-shelf solution, with the object as the priority, not the user. We have shown that the structure may actually be manufactured in the same way, but the difference is the effort and priority you put into the design.”
What will the future look like?
The UK government wants the construction industry to shift towards manufacturing and digital processes, believing this will deliver better value to society. UK Research and Innovation’s Sam Stacey says: “Longer term, the vision for the built environment is that we define a ‘platform rulebook’ for different asset classes. We identify the parts that are required to make these assets, the performance criteria for these parts and the interfaces between them.”
He gives the example of a typical three-pin plug and socket: “There are myriad ways that you can make these things, but the pins and holes have to be the same and there are strict performance criteria in terms of volts and amps.”
The same thing could be done for bridges or buildings to create a ‘kit of parts system’ for each asset. “What we want is for lots of different manufacturers to be making parts for the system, and we want there to be innovation for the design and manufacture of these parts and the economies of scale that come with all that,” he says.
The UK government wants the construction industry to shift towards manufacturing and digital processes, believing this will deliver better value to society
When this ‘kit of parts’ has been developed for every asset, Stacey expects designers to adopt a ‘generative design’ approach. “Once we’ve got a set of standardised rules, you can put in the particular characteristics of the project and the AI works out the design,” he explains. “If the system is working to its maximum, it will be vastly better than what we have today. We will be constructing built assets with a fraction of the labour, and then you start to introduce automation, robotics and so on and the sky’s the limit.” Other benefits include cutting material waste, reducing over-design and improving maintainability.
Stacey says this approach should not limit designers’ creativity: “There are a lot of examples of products we admire that are quite standardised – for example in the automotive sectors. A certain amount of tweaking of the design is inevitable, because of different ground conditions and so on, so you standardise 80% of the product and the final bit you do specially for the site. You get the AI and automation to do the grunt work, and you apply creativity to make it the perfect fit for the location. You can get more value from your creativity if you take this approach.”
Sign up to receive news from ICE Knowledge direct to your inbox.