From lab to fab: how AI is closing the gap between materials discovery and manufacture
Advanced materials innovation focuses on engineering materials with tailored properties, from improved strength and conductivity to reduced weight and enhanced sustainability. What can AI do to help?
By Sean Hargrave
Introduction
Materials set the parameters of innovation. AI brings hope for change
Every industrial product inherits the limits of its materials. An aircraft is only as light as its alloys and composites allow. A data centre is only as efficient as the semiconductors managing its power. An electric car costs what its battery chemistry costs. Before any engineer touches a product design, the materials have already set the parameters.
This can be restricting. Time has always been a challenge in materials discovery and testing. A new material typically takes 10 to 20 years to move from discovery to market. Lithium-ion batteries, for example, were proposed in the mid-1970s and did not reach mass adoption until the late 1990s.
And yet AI is already changing expectations. In November 2023 Google DeepMind published its GNoME model in Nature, reporting 2.2 million predicted crystal structures, of which 380,000 were judged stable enough to pursue. Confidence in the sector is growing. Recent VC funding deals have gone some way to illustrate this, with Periodic Labs, founded by former DeepMind and OpenAI researchers to automate materials experiments, raising a $300m seed round in 2025. More recently, UK start-up CuspAI raised $450m (£338m) in funding from investors including Bezos Expeditions and the UK Sovereign AI Venture Fund.
The industries waiting on faster materials are not niche. Airbus’s push into AI-assisted design is aimed squarely at the airframe, to cut weight with new composites. Semiconductors need materials that handle higher power with less waste heat, which is why CuspAI says 80% of its research this year is going into chip materials. Batteries, medical implants, wind turbine blades, and satellite components all sit in the same space, where a material has to be discovered, tested, and proven safe before engineers can design around it.
The link between AI and university research is already producing results. Materials Nexus, a London start-up, used machine learning to predict an alloy composition with the magnetic properties of a rare-earth magnet, without the rare earths. The Henry Royce Institute’s facility at the University of Sheffield produced and validated a physical sample. Materials Nexus has since won £700,000 in Innovate UK funding to refine the model and search for further rare-earth-free alloys.
The UK has built its policy around this kind of company. Over 2,700 businesses are active in materials innovation, 90% of them SMEs, and 74% employ fewer than 50 people. These are organisations without the balance sheet to wait a decade for a return on investment in materials discovery. Materials-specific roles in these companies already contribute up to £4.4bn a year to the economy and the National Materials Innovation Strategy, published in January 2025, sets out an ambition to at least double the number of these jobs by 2035.
What follows in this AI and advanced materials briefing paper is an examination of where AI is actually closing that gap, and where validation, scale-up, and certification still set the pace.
A digital publication from Bristol Innovations, University of Bristol
Section 1
How AI is transforming advanced materials across key markets
Materials domain
Key markets driving demand
How AI changes development and production
What constrains scale-up
Semiconductor materials & advanced packaging
AI compute, defence systems, telecoms, automotive
Improves chip design, materials optimisation, defect detection and yield in fabs
Capital intensity, supply-chain concentration, materials bottlenecks, energy use
Advanced composites
Aerospace and defence, robotics, transport, wind energy
Improves simulation, structural optimisation, layup accuracy and defect detection
Certification cycles, cost, repeatability, integration into existing production
Links materials properties with process parameters and quality outcomes
Data fragmentation, interoperability issues, resistance to process change
Sustainable and low-carbon materials
Construction, infrastructure, public sector procurement
Balances performance, cost and environmental impact in materials design
Verification of sustainability claims, standards alignment, adoption risk
Sources: Deloitte; Omdia; OECD; World Economic Forum; National Materials Innovation Strategy; BI Foresight analysis.
AI-materials discovery funding, 2024–2026
Venture money has moved fast, and it has moved on both sides of the Atlantic. CuspAI, founded in Cambridge in 2024, raised $30m at seed; by September 2025 it had added a $100m round at a $520m valuation; in July 2026 it closed a $450m round at $2.6bn – a fivefold increase in ten months. Periodic Labs, founded by former OpenAI and DeepMind researchers, raised $300m in a single seed round in September 2025, valuing it at $1.3bn. Radical AI, building autonomous materials labs in New York, raised $55m at seed the same month. Two Imperial College London and Munich-based spinouts – Polaron and ExoMatter – show the same pattern at earlier stages. All five companies build AI tools for materials discovery or design specifically, not general-purpose models with materials as one application among many.
CuspAI
Seed · Jun 2024 · Cambridge, UK
$30mundisclosed
Radical AI
Seed · Jul 2025 · New York, US
$55mundisclosed
Periodic Labs
Seed · Sep 2025 · San Francisco, US
$300m$1.3bn–$1.5bn valuation
CuspAI
Series A · Sep 2025 · Cambridge, UK
$100m$520m valuation
CuspAI
Series B · Jul 2026 · Cambridge, UK
$450m£338m · $2.6bn valuation
$0$450m
ExoMatter
Pre-seed · Oct 2024 · Munich, DE
€1.7m
Polaron
Seed · Feb 2026 · London, UK · Imperial College London spinout
$8m£6m
$0$8m
According to Bloomberg, Periodic Labs is now in advanced funding discussions to raise at least $500 million at a valuation of approximately $7.0 billion to $7.5 billion.
Sources: CuspAI: SiliconANGLE, TechCrunch, Bloomberg (via TechFundingNews), Reuters (via Yahoo Finance), 20 July 2026. Periodic Labs: TechCrunch, 30 September 2025; Contrary Research, June 2026. Radical AI: Pulse2, FinSMEs, Dealroom, 20–21 July 2025. Polaron: Imperial College London, EU-Startups, February 2026. ExoMatter: EU-Startups, Tech.eu, ZAKA VC, 24 October 2024.
Section 2
Development landscape – modelling and simulation
AI speeds up discovery
Bassam El Said, senior lecturer in Digital Design and Manufacture of Composites at the Bristol Composites Institute, works at the forefront of these new demands. His focus is on using AI to help teams of researchers discover new materials for aeroplanes and wind turbines that are lighter and more resilient to temperature extremes and vibration, so they can deliver longer range and greater efficiency with lower emissions and at a more affordable cost.
AI is empowering El Said and his fellow researchers to achieve in hours what would previously have taken months.
“Materials are very complex because there are so many options when you are mixing layers of carbon fibre, glass fibre and ceramics and you can have a wide variety of individual fibre length and width,” he says. “There are millions of possible combination choices, you just don’t have time to test every one.”
“It’s like having hundreds of PhD students running lab experiments all at the same time.”
Bassam El Said
Senior lecturer in Digital Design and Manufacture of Composites, Bristol Composites Institute
Traditionally, adds El Said, researchers would test what they think are the best options in the lab but it would still take a long time. That was until AI came along.
“We’re now able to train AI to understand each of the components so it can simulate multiple combinations,” says El Said. “It’s like having hundreds of PhD students running lab experiments all at the same time. It doesn’t replace the human researcher, but it does allow them to run many simulations so they can identify a new material worthy of more research in hours rather than months.”
While it’s making an impact in research, El Said predicts it will be another decade yet until today’s new AI-inspired material discoveries will make their way from the lab and into the planes, satellites, and clean energy infrastructure of tomorrow. That is because whatever materials researchers design will have to go through rigorous testing in manufacturing scenarios because there are many variables in how they turn out in real life rather than in the laboratory.
AI’s push into the final frontier
It’s at this interface of lab and industry where the UK Space Agency reports some of the most interesting new materials work is taking place. It’s hard to think of an environment that is more demanding on the performance of materials than orbiting the planet. The increasing demand for satellites is prompting a search for lighter, stronger materials that can withstand high radiation and extreme temperature cycles while also reflect less light, so dark skies astronomy is not hindered.
It’s a tough list of requirements but, according to Dale Wyllie, senior payload systems engineer, UK Space Agency, AI is greatly speeding up the testing of new materials at low costs. This is particularly true in running simulations in which researchers explore molecular dynamics, predicting how atoms and molecules interact over time in harsh environments.
“A good example is high-entropy alloys – materials made from multiple elements that show real promise for surviving the harshness of space,” he says. “Because of their complexity, AI and ML are essential tools for understanding how these alloys behave, dramatically shortening what would otherwise be a lengthy design and testing cycle.
“NASA is exploring how AI can work alongside additive manufacturing to rapidly prototype structures that are optimised to handle the stresses of launch.”
Dale Wyllie
Senior payload systems engineer, UK Space Agency
“AI is also being applied to structural design. NASA is exploring how AI can work alongside additive manufacturing to rapidly prototype structures that are optimised to handle the stresses of launch – moving from concept to physical test piece far faster than before.”
Wyllie believes the technology is advanced but vital data often sits in siloes, rather than being shared. While he is hopeful the government’s AI for Science Strategy might help release some high value datasets so they can be shared by researchers, digital twins are the main area to watch in the meantime.
“Alongside data, digital twinning is likely to be one of the next major steps forward,” he says.
“By combining real-time computational modelling with rapid physical prototyping, digital twins could allow AI to continuously refine its understanding of how new materials and structures actually perform – bridging the gap between the lab and the factory floor that has historically slowed the journey from discovery to deployment.”
Future of manufacturing
If AI is to have real-world impact, the modelling and digital twin work will need to move beyond the theoretical into practical improvements to processes on the manufacturing shop floor. It is one thing to design a new material or production system, quite another to create a new way of working that can be put to good use.
So, if observers are looking for an early glimpse of how AI can help shape the future of manufacturing, it would be worth looking at the work being carried out at the NCC (National Composites Centre). Its researchers are demonstrating a new model for making liquid resin infusions. This process to make advanced composites is beset with potential problems that can be hard to rectify if they are not acted on immediately.
This is where the centre has used advanced knowledge from engineers to train an AI model to boost efficiency by monitoring the infusion process and spotting anomalies requiring correction. The centre sees this as the perfect combination of human know-how being trained into AI, and gives a practical example of AI aiding materials manufacturing on the shop floor.
“We’re excited about AI for liquid resin infusion because, despite the buzz around AI, it can be difficult to find examples of it making a difference in real manufacturing,” says Dan Griffin, principal research engineer for Automation and Digital Systems at the NCC.
“For complex manufacturing processes, AI can’t learn what it needs to by scraping data. We need human engineers to build bespoke environments in which to train machine learning models. The resin infusion work is the result of exactly that – and it works. Machine learning models control the resin infusion process in real time, spotting defects before they become critical and adjusting the process in response. We’ve already demonstrated it on live infusions in controlled conditions – and now we want to test it in an industrial setting. AI can contribute to a more intelligent, adaptive, competitive and sustainable industrial base – but only with the right combination of process, collaboration, and applied engineering.”
AI’s first impact, then, is not so likely to be a completely new material breakthrough, but rather a way for engineers to better manage and optimise existing cutting edge manufacturing systems – a means to handle trade-offs on the factory floor before we see a revolutionary material introduced.
Two spinouts using AI to advance innovation
Two of the University of Bristol’s own spinouts show the same pattern the rest of this briefing describes, without any AI in the technology itself. ICOMAT, founded by University of Bristol researchers in 2019, makes carbon-fibre composites using a fibre-steering process that avoids the defects standard manufacturing produces. It raised a $22.5m Series A in 2024, backed by 8VC and the NATO Innovation Fund, and has since taken £4.8m from the UK Space Agency to build a Gloucester factory serving aerospace, defence, and space customers.
Anaphite, also founded by Bristol graduates, makes a dry-coating process for lithium-ion battery electrodes that cuts manufacturing energy use by around 30%. It has raised more than $20m across several rounds, including a $13.7m Series A in 2024.
Neither company uses AI as its core technology. Both are evidence that a research base capable of producing Materials Nexus and the Bristol Composites Institute’s AI-assisted composite work is also capable of producing investable materials companies without it, a reminder that AI is just one route to commercialising UK materials research, not the only one.
The Bristol Innovations Zone connects businesses with world-leading researchers to turn R&D ambitions into real outcomes. Find out how.
Spanning the evident gap between research and real-life use will be by far the biggest step for any company working with AI and new materials to address. It’s in spanning this gap that businesses will identify the need for a regulatory framework that is up to date as well as a work force equipped with the necessary skills.
Airbus is at the forefront of high-performance engineering and reports huge excitement offered by the potential for AI to help develop new materials. These should allow it to make lighter airframes that are robust enough to perform well under severe stress, including wide ranges in temperature, pressure and wind speed.
The aircraft manufacturer believes AI holds considerable promise in hybrid testing of materials, where computer simulations run alongside physical testing. The models are now sophisticated enough to play a crucial role in speeding up discovery, but there are some very practical steps that will need to be in place before AI spans the gap between aiding new materials discovery and testing and then real-world manufacturing.
“The biggest challenge will be certification, we need a regulatory environment that moves at the pace of technology,” says Pooja Narayan, fast track lead for artificial intelligence at Airbus.
“Airbus is currently working with regulators to establish the first standards for implementing AI onboard. To make this vision a reality, regulators must upskill alongside industry to identify gaps and implement evidence-based solutions.”
In fact, retraining is required across industries and regulators before AI delivers its full promise, and that promise will be far-reaching, encompassing a cradle-to-grave approach for new materials.
“The biggest challenge will be certification, we need a regulatory environment that moves at the pace of technology.”
Pooja Narayan
Fast track lead for artificial intelligence, Airbus
“AI may be integrated across the entire lifecycle, from design and manufacturing to embedded services. This isn’t just about software; it’s about the industrialisation of these technologies at scale,” Narayan says.
“The practical implementation requires a wide upskilling of both staff and regulators to manage these new non-deterministic systems. The primary challenge is ensuring that as we move toward AI-driven workflows, we maintain the gold standard of safety the public expects.”
While these desires to modernise regulation frameworks and upskill a new generation of workers are lofty ambitions, there is a more mundane challenge to the rollout of AI and new materials.
Nina Gryf, senior policy maker at Make UK, offers a cautionary note that, while the manufacturing group’s members see AI as holding huge potential, a massive 99% fit in the bracket of SMEs. This suggests they don’t have the time, headcount or capital to run ambitious AI research programmes alongside day-to-day production. That means the AI work currently being carried out by the country’s manufacturers is, in her experience, focused on quicker, more contained wins, such as predictive maintenance and analysing energy use patterns, rather than materials discovery work.
Section 4
Q&A: Professor David Knowles, CEO, Henry Royce Institute
Professor David Knowles
CEO, Henry Royce Institute
Q. Why is there currently such a huge focus on materials?
Materials sit at the heart of both the country’s new industrial strategies and huge opportunities are being seen with the link to the digital transformation and AI. That’s why the industry’s being called Materials 4.0 to match the Industry 4.0 that reflects how digital now runs through the core of every material’s life from discovery to manufacture and from use to recycling.
It’s a very exciting time to be in the field as academia and industry need to work together because materials are at the heart of everything. You can’t have quantum computing without cryogenic and quantum materials, as they’re working on at Imperial in London, Cambridge, and Manchester, you can’t have clean energy without battery improvement, such as they’re working on at Oxford University, and you can’t make lighter, more environmentally-friendly cars and planes without new metals and alloys, as they’re researching at Sheffield. There’s work going on across the country.
Q. How is AI being deployed?
AI is greatly speeding up discovery because it can analyse experimental and synthetic data in vast quantities to probe the capabilities of new materials. It can also oversee and inform automated robotics tests on material samples which would otherwise take many researchers several months. This high throughput is very useful because it gives prompt feedback for further testing. You can set a parameter or objective, such as a material being 10% more efficient, and it will keep on feeding back on that objective.
It’s also very good at helping us to see which data is important. Research and testing can generate so much data that you need to know where to focus.
Q. Where is AI struggling?
Access to trusted data is the key. Data is always king but so much of it is not accessible because it’s from research work that gets stored away or poorly curated. Also, much of what a company is doing relates to know-how and so they keep it locked away. Nobody’s quite sure what the value of their data is and so they don’t always want to share it.
It’s completely understandable, we just need to find a way to unlock it across multiple applications. Part of the solution would be a better way of acknowledging where the data has come from. We have very good tools to search through papers, but it can only find publicly available data.
Q. How close are we to AI-discovered new materials?
We’re on the cusp of some exciting developments in discovering metamaterials that perform better and are more environmentally friendly. The government is supportive and so we are expecting Innovate UK to get behind some interesting initiatives with funding to help us discover and test new materials but then go beyond the gram level to being able to scale up and manufacture. That’s the gap we need to span. To go beyond developing to see advances in basic materials chemistry knowledge being translated into manufacturing.
Section 5
Perspectives
Semiconductor industry VC – Lawrence Lundy-Bryan, partner, Cloudberry
Cloudberry VC is Europe’s first dedicated semiconductor venture capital firm. Headquartered in Helsinki and London, the fund backs early-stage (pre-seed and seed) deep-tech start-ups. It invests between €400,000 and €1,000,000 in European companies developing foundational hardware, specifically focusing on semiconductors, photonics, and advanced materials.
Is AI-driven materials development on a VC’s radar?
Very much so. Periodic Labs raised $300m at a $1.5bn valuation last year, led by a16z with Nvidia, OpenAI, and Bezos. Former OpenAI and DeepMind researchers building autonomous labs for materials discovery. So the category is clearly hot.
Where we sit is slightly different. What excites us isn’t so much discovering entirely new materials, it’s using AI and new manufacturing tools to optimise how existing materials are combined and processed. In semiconductors, the bottleneck isn’t “we need a material that doesn’t exist yet.”
Which markets are really calling out for this?
Thermal management in semiconductors is calling loudest. As chips get more powerful and more densely packed, heat dissipation becomes the limiting factor. We’re investing in a company working on diamond thin film deposition for exactly this reason – not a new material, but an innovative approach to combining materials that has immediate commercial relevance because the fabs need it now.
What concerns would you need answered before investing?
The lab-to-fab gap is the big one. It’s one thing to demonstrate a promising material combination in a research setting, it’s another to manufacture it at scale with consistent quality and at a price point the industry will accept. Foundry compatibility is critical: semiconductor fabs are incredibly sensitive environments, and any new material has to prove it won’t contaminate existing process lines. That’s a genuine barrier, not a technicality. We’d want to see a clear path to production, ideally with manufacturing partnerships already in discussion.
Sustainability VC – Stephen Price, investment partner, Clean Growth Fund
The UK Clean Growth Fund is a specialist climate technology venture capital fund that invests in early-stage UK companies developing innovations to cut greenhouse gas emissions and improve resource efficiency. Backed by private investors and the UK government, it provides critical seed and Series A funding (typically £500k to £5m) to help clean-tech start-ups scale their solutions and reach the market.
Is the development of new materials developed by AI a promising area that excites VC?
Yes! Deep tech is gaining a share of VC investment, and advanced materials is a component within that. It’s an intriguing technology area across multiple industries and application areas. For climate tech I’d call out batteries and energy storage, lightweight aviation materials, thermal management, and advanced processing for critical minerals and rare earths, particularly around catalysts, membranes, and sorbents.
What are the concerns you need answered before investing?
Scalability. AI materials discovery is one thing but manufacturability is quite another. Innovators would need to show how they can scale processing of advanced materials and what that would cost. Industrial partnerships are likely to be key.
A poster child always helps but it’ll come down to the fundamental performance vs cost equation. VCs will be attracted where there is a compelling value proposition, solving urgent problems in a large, growing market – and where innovators can get to scale in as capital-light a way as possible.
Academia – Professor Martin Kuball, head, Centre for Device Thermography and Reliability at the University of Bristol, and chair in Emerging Technologies, Royal Academy of Engineering
Martin Kuball leads the £5m EPSRC Programme Grant GaN-DaME, which develops GaN-on-Diamond technology for ultra high power RF devices, and the £2m EPSRC Platform grant MANGI, which implements this technology for next-generation internet applications.
What is most likely to be AI’s main early success in developing new materials?
The area where I see people making the biggest inroads is the efficiency of the power unit that compound semiconductors can enable. This is where silicon can be improved on for efficiency and cooling. Rather than using water and fans to take away all that heat.
What will be the impact of AI creating new materials?
New designs created by AI and possibly new non-silicon materials discovery by AI will help us produce chips that use materials that can work at higher voltages, higher efficiency, and have better heat management capabilities. There’s a lot of energy and heat released every time the electricity from the grid is stepped down to work with chips at low voltage. So, reducing the number of semiconductor components to stop down this voltage, which is possible using so-called or ultrawide bandgap compound semiconductors, would be more efficient and environmentally friendly.
Industry – Dale Wyllie, senior payload systems engineer, UK Space Agency
The UK Space Agency (UKSA) is the government department responsible for the United Kingdom’s civil space programme. Part of the Department for Business, Innovation, Science and Trade, it directs national space strategy, funds scientific research, licenses launches, and manages partnerships with international bodies like the European Space Agency
How is AI helping to answer these challenges in design?
The use of AI in engineering is still in its infancy, and engineers are still finding novel ways to incorporate AI methods and tools into the design process across the mission lifecycle to improve efficiency and allow more extensive trade-offs. For example, AI is accelerating materials discovery in ways that would simply take too long using traditional methods. One key technique is molecular dynamics – simulating how atoms and molecules interact over time – which gives researchers a window into how a material might behave in the space environment before it has ever been made. Machine learning is enhancing this further through ML interatomic potentials, which can model complex materials far more quickly and cheaply than conventional approaches, striking a practical balance between computation time, cost, and accuracy.
What is the missing link, the next advancement needed in AI and materials design?
The UK Space Agency is always interested in understanding new applications of emerging technical approaches and methodologies, including the adoption of AI into the engineering toolset. High-quality training data is arguably one of the most critical gaps right now. AI is only as good as the data it learns from, and in materials science that data is often scarce, siloed, or inconsistently formatted. DSIT has recognised this directly, with a commitment in the AI for Science Strategy to identify and develop high-value datasets that can unlock breakthroughs in priority areas like materials discovery.
Industry – Emre Ozer, senior director of processor development, Pragmatic Semiconductor
Pragmatic Semiconductor is a UK-based deep-tech company that designs and manufactures ultra-thin, flexible microchips (FlexICs). Unlike rigid silicon chips, their proprietary technology uses thin-film transistors on a flexible substrate, which allows intelligence and connectivity to be embedded seamlessly into everyday items at a significantly lower cost.
Can you tell us about your work with AI to develop new materials?
We’re in the process of setting up some research which will use AI to help us generate the material for the next generation of our flexible chips so we can shrink transistors even smaller. We already work with clients who use AI tools to design the chips they want us to make for them. The next step is our work to use AI to allow us to discover new materials a lot faster, so we can design efficient large-scale flexible integrated circuits.
What impact can AI make?
Nanometre precision is incredibly difficult, and the yield challenges are where AI can help most. More broadly, anything that improves performance-per-watt, whether that’s new substrate combinations, advanced packaging materials, or better thermal interface solutions, has a clear buyer in the semiconductor supply chain today.
How do you see AI fitting in with your R&D efforts going forwards?
With AI, we could narrow down our R&D focus to, say, two new materials and then our engineers will go to the lab to fine-tune them and make the best option ready for scalability. If we don’t use an AI tool, that means we need to maybe assign ten R&D engineers. They all explore different materials themselves, theoretically as well as in the lab. It will take years, several years. We will shorten that time using AI tools. But still, we need researchers to go to lab and do the tests.
Section 6
What to watch
1
Whether digital twins move from proposal to practice.
Dale Wyllie names digital twinning (combining real-time computational modelling with physical prototyping) as the next major step after data access. No UK space or materials programme has yet demonstrated this at scale; the first working example will be the signal that the lab-to-factory gap is genuinely closing rather than just narrowing.
2
Whether the National Materials Innovation Strategy’s steering groups actually convene.
The strategy names Materials 4.0 as one of two priority cross-cutting themes and commits to a dedicated steering group for it. Whether that group is resourced and meets on schedule, rather than existing only on the implementation diagram, will show whether the strategy is delivery or intention.
3
Whether Innovate UK’s 2026 funding competitions get taken up by materials specifically.
The Frontier AI Benchmarking Datasets competition (up to £4.5m) is designed to fund exactly the curated, shared datasets that Knowles and Wyllie both identify as the missing link. Whether materials science wins a meaningful share of it, against biosciences and semiconductors, is worth tracking.
4
Whether certification keeps pace with the technology.
Narayan’s warning that regulators must upskill alongside industry is currently just a warning. Airbus’s work with regulators on the first standards for onboard AI is the concrete test case; a published standard, or a stated delay, will show whether that keeps pace or becomes the bottleneck she predicts.
5
Whether SME funding for materials AI actually increases.
Gryf’s account of Make UK’s membership (99% SMEs, with AI use limited to predictive maintenance rather than materials discovery) is a funding-access problem, not a capability one. A meaningful shift up that ladder, into discovery work, would be the clearest sign the strategy’s ambition to double materials jobs by 2035 is reaching the businesses that most need it.
6
Whether the machinery of government holds.
The National Materials Innovation Strategy and the AI for Science Strategy were both developed under DSIT, now abolished and folded into the new Department for Business, Innovation, Science and Trade. Whether responsibility for these strategies, and the ministers who championed them, carries over intact is unresolved as this briefing goes live.
7
Whether NCC’s resin infusion model reaches an industrial setting.
Griffin is explicit that the technology has so far only been demonstrated on live infusions in controlled conditions, and that the next test is an industrial one. That move from controlled demonstration to a working factory floor is the single clearest test of this briefing’s central argument: that AI has already solved discovery, and deployment is where the real work now sits.
Active in deep-tech R&D?
The Bristol Innovations Zone (BIZ) is one of the UK’s leading gateways for deep-tech collaboration – connecting businesses with University of Bristol researchers, facilities and the wider innovation ecosystem.
