Recent research has shown that manufacturing leaders are keen to make start-up collaboration part of their business strategies. The Manufacturer’s Joe Bush reports.
A survey of 100 UK manufacturing leaders, conducted by Digital Catapult at the end of last year, showed that manufacturers are increasingly looking to partner with start-ups to stimulate innovation, with nearly three quarters of manufacturing leaders (73%) saying that collaborating with start-ups is part of their long-term business strategy. This will help to accelerate digital transformation objectives by tapping into skills and creativity that are perhaps not available in-house.
These survey results came off the back of the conclusion of Digital Catapult’s 12-month Made Smarter Technology Accelerator, which resulted in start-up-led minimum viable products (MVPs) developed in record time for industry giants such as BAE Systems, Babcock International Group, Northumbrian Water Group and Safran Landing Systems. Jeremy Silver, Digital Catapult CEO, described the programme as: “Hard evidence of what can be achieved by open innovation programmes – successful, real world solutions delivered against the clock to drive real change for manufacturing giants.” He did add a warning however, that despite the positives (and perhaps part of a risk-averse COVID recovery), many manufacturing heavyweights are not making collaboration with innovative early stage companies a priority in the next year. As a result the survey suggests that there is a danger that UK companies will miss out on the opportunity to be early adopters of new technologies, and learn from start-ups’ creativity, innovation and agility.
The Made Smarter Accelerator
As part of the accelerator programme, two start-ups, Riscon Solutions and Inventia UK, joined together to work with Northumbrian Water (NWL) on the water network monitoring and real time challenge. The ITERATION project delivers a cloud-enabled, remote water quality monitoring solution, which allows Northumbrian Water to improve its visibility and understanding of water quality within a designated district meter area (DMA). Water has several stages of treatment until it reaches homes as clean drinking water, and during this supply chain there are several problems which may occur such as leaks and contamination.
The way these issues have traditionally been mitigated is via manual sampling – a very expensive and labour-intensive process. “Within the water industry this process has been improved somewhat by using data loggers,” says Amir Kotb, Head of Technology and Founder of Inventia. “What we have produced is a state-of-the-art revolution of that. We have used narrowband IoT technology provided by network operators. And we have designed, manufactured and integrated a hydraulic system to take samples from the water, and put it in a specific pressure and flow that is suitable for the sensors. This helps monitor and manage water quality, flow and pressure in near real-time.”
Roger Singleton, Managing Director of Riscon Solutions adds: “Thus far an MVP prototype is currently gathering water quality data from the main water lines every 15 minutes and transmitting this remotely up to a custom software dashboard. This will help identify and notify potential and actual leaks within the system. “NWL clearly identified that this monitoring is a gap in their armoury and recognised the need for a digital solution in an engineering space. Riscon’s partnership with Inventia has the right skill sets to complete the challenge, as it is a multi-disciplinary problem.
I don’t think a traditional IT company could have done it, as it needed a huge amount of engineering knowledge combined with digital skills.” Machine Intelligence, a developer of computer vision and machine learning techniques, is working with BAE Systems on a scalable artificial intelligence for visual inspection challenge, using a novel machine learning technique. A prototype is identifying defects in manufacturing, improving the quality of the end product, reducing waste and saving money. Simon Harding, Director at Machine Intelligence, explained that the company has its own AI stack which is not based on technology from elsewhere. “This technology works in a fundamentally different way to other solutions that you may come across,” he says. “Using evolutionary algorithms, it generates computer programmes which can be understood by a human, ported to different platforms, edited, debugged and tested.
And where this really helps manufacturers is when they have very few training examples. “One area where we’ve be working with BAE is that inside aircraft, there are numerous pipe sections that carry a variety of inspection challenges. We picked an X-ray inspection of pipe welds and developed a system that sits next to their X-ray machine. They can X-ray the pipe welds using an automated robotic X-ray scanner. The images then automatically flow into our software, where they’re analysed and sentenced. The operator gets to see where the AI has picked up defects or features of interest.
“This job was traditionally done by a human who had to spend hours looking at pixel level through piles of X-ray images. This was extremely time consuming, costly, and open to mistakes and operator fatigue.” Working with Babcock International Group on the warrior base overhaul challenge, JetSoft, a provider of solutions to improve the capture, management and utilisation of inspection and quality control information, is developing a solution that will utilise existing equipment and combined data output to create an automatic 3D dataset – easing the hull inspection process.
Existing solutions target improved traceability relying on the use of automated systems. However, these systems perform poorly on some components’ geometries and require large investment by the users. The proposed system will integrate existing technologies and working practices, requiring less investment and retraining. Producing rich 3D datasets, to better understand about equipment and health trends, the solution will improve inspection operations and enable detailed analysis of advanced equipment and components.
“In our view, inspection data is some of the most valuable data there is in manufacturing,” says Jetsoft CEO Tom Martin. “If you’re looking to reduce defects, improve quality and reduce waste, inspection data is the core information that enables this to happen.” Jetsoft’s initial contact with Babcock was around combining positional and inspection data, to create improved traceability. “However, it became clear when we were on-site that if we achieved that, that data would have to go somewhere.” Martin adds. “They also needed a system to curate this data and generally improve their digitisation. We took on that challenge and expanded the brief to look at how we could digitise their processes. Babcock was incredibly helpful and involved with the project. They supported what we were doing, were responsive and great to work with.”
A word from the start-ups
What are the benefits of working with start-ups?
Speaking to the start-ups involved in the Made Smarter Technology Accelerator it is clear that their size, responsiveness and flexibility are a key selling point when dealing with larger entities – making innovation quicker and easier, thus easing the route to market for new ideas and concepts.
“UK manufacturers should work with different sizes and stages of companies depending on their needs, but start-ups are best if disruption is needed – to shake things up,” says Singleton. “However, bigger companies need to understand this is a journey, and if start-ups achieve the goals set to them, they’re going to need customers to survive.” Harding adds: “Big organisations like the responsiveness and the flexibility that smaller companies can give them. We started off the project with BAE trying to solve a particular engineering problem. What we’ve ended up with is a complete system, which fits their exact process needs, and can dovetail very nicely with their existing method of working. This is something that I don’t think a big company would be able to do within the timescales or within the budgetary boundaries. We can just be more flexible.” Start-up and small companies tend to be innovative by nature because they need to be. And in order to survive they must be doing something new and different. Therefore, start-ups have the freedom to be agile and implement solutions quickly. “JetSoft is knowledgeable on what’s achievable with inspection data, which isn’t necessarily the case with internal people in manufacturing. Having wide industry knowledge allowed us to develop solutions that are commercially beneficial,” adds Martin.
Are start-up companies an untapped resource for UK manufacturers?
There are more start-ups launched per year in the UK than in the rest of Europe combined, especially around AI. So, there’s certainly the skills available around machine learning, and the various technologies that go alongside it, and they’re ripe for exploitation. In addition, as Harding explains, it is a misconception to think that all this talent is concentrated in London and being worked on in financial districts, which is the normal draw for AI technology.
“We’re based in a tiny village in Somerset – there are lots of skills all over the country and companies can essentially find someone local to them who can help solve their problems.” Martin adds: “Manufacturers aren’t particularly good at identifying how they can improve, whereas start-up companies have some great ideas. A lot of the work that JetSoft does is educating the manufacturing market on what’s possible. Start-ups have evolved due to finding a gap in the market or developing a niche way to improve wider processes.”
“Something like the Made Smarter programmes highlight how start-ups really can benefit manufacturers. They bring new ideas and concepts to the market. So certainly, they should be utilised more,” says Kotb. In order for collaborations to be a win-win for both parties, Singleton would like to see more longer-term thinking by the bigger industry players to see how engaging with an existing start-up could lead them to develop into a much larger offering. He stressed, however, that this will need more programmatic style thinking (five-to-ten year development journeys mapped out with business benefits and technical milestones) around R&D in the bigger players in industry. “It is a current skill gap in my opinion. I also feel the US do this much better than the UK at present. They have a much bigger appetite for longer-term investment,” he says.
Do manufacturers have a perceived risk around working with start-ups?
Of course, some of the benefits of being a small and agile company can also serve as a drawback. Quite a large number of start-up companies cease trading for various reasons, whether they get bought up or are not financially successful. And that always introduces a risk – particularly for big companies who are seeking long-term relationships.
And, while start-ups are great problem solvers, they are generally light on production and customer service capacity. “This is where programmes like Made Smarter can come in,” says Harding. “They help take off a lot of the financial burden – these government grants are great for that. They help out with de-risking that initial investment. And of course, they make the SME more likely to succeed in future.” Singleton adds: “Taking products from TRL6 (technology readiness level) through to TRL9 is often known as the valley of death. I think this will only be mitigated with investment to allow companies to get their products match fit.
The industry needs to think more about how to support that better moving forward, otherwise a lot of ideas will remain as nice proof of concepts.” “I believe the concerns come, not from a worry about the product or service that a small company would provide, but more around the longevity of the company and lack of trading history,” adds Martin. “The adage ‘nobody got fired for buying IBM’ comes to mind and I think this risk is overplayed. “Manufacturers can insulate themselves from it by managing projects closely and working in collaboration with suppliers to help define scope and detailing project milestones. The risk/reward is heavily in favour of using start-ups who normally revolve around a committed team who work hard and are keen to succeed. I would like to flip the IBM quote around and say nobody excels by doing the same as everyone else.”
Key takeaways
- 73% of manufacturing leaders are making collaboration with start-ups part of their long-term business strategy
- Start-ups are inherently creative, innovative and agile
- Perceived risks of working with start-ups, from risk averse manufacturers, need to be mitigated if opportunities are to be realised
- Start-ups have the freedom to implement solutions quickly
- In the UK there are many skills around machine learning and related technologies that are ripe for exploitation
For more information visit Made Smarter
For related articles visit this section of our website