Why robots are being locked out of British factories. By Nick Thompson

Three quarters of British manufacturing SMEs operate without a single robot, according to Make UK, despite a decade of government grants, skills programs and financing schemes designed to change that. The International Federation of Robotics puts the UK’s robot density at 112 per 10,000 manufacturing workers – roughly half the EU average and well below the global average of 151. That picture has barely shifted despite years of intervention.

The received wisdom is that the problem is money. Manufacturers can’t afford robots, so the response is to subsidize hardware, fund pilots and make financing more accessible. I’ve had this conversation many times. Money matters, but in my experience, it is not the only, or even the decisive, constraint. What consistently isn’t available are the people who know how to get a robot actually working.

The specialist bottleneck

Getting a robot operational is not primarily a hardware problem. The hardware – a cobot arm, an autonomous mobile robot, a vision-guided inspection system – is a mature, proven product. The hard part is everything that comes after it arrives. Programming the robot. Integrating it with existing systems. Validating that it does what you need in your specific environment. That work often requires specialist robotics engineers who understand ROS2, the widely adopted open-source robotics ecosystem used to build, integrate and test modern robotic systems.

These specialists are in genuinely short supply. For a mid-market manufacturer with an automation opportunity, the practical options are challenging. Engage a systems integrator and commit to a bespoke design-and-build process at significant upfront cost, with no validated outcome until late in the project. Or shelve the idea.

Most are choosing to shelve it.

The people who understand the problem – the production manager who knows exactly where automation would help, the process engineer who has thought through the workflow – are already in those organizations. They are not the gap. The gap is that the infrastructure for robotics development was built for someone else.

A software problem in an engineering wrapper

Many manufacturers already have access to software capability, whether in-house, IT teams, or through trusted development partners. What those developers have never been able to do is apply those skills to a robotics problem, because the toolchain wasn’t designed for them.

I spent two decades building software teams. I watched what happened when cloud platforms arrived, not just to infrastructure costs, but to who could build things. Developers who had no interest in managing physical servers could suddenly deploy sophisticated systems because the complexity had been abstracted away. The underlying technology didn’t simplify. The access did. That shift produced more capable engineers, not fewer, because more people could participate. Robotics is at the same point. The frameworks exist. What’s been missing is the abstraction layer that lets the people who understand the problem get their hands on it.

Platforms built on top of ROS2 now offer browser-based environments, simulation, and SDKs in languages that mainstream developers already know. A developer can model a production scenario, test behaviors in simulation and de-risk an application before any hardware is ordered.

The economics of experimentation

The traditional path to automation front-loads the risk. You engage specialists, commit to hardware, run the integration and then discover, over months, whether the outcome meets expectations. The Manufacturing Technology Centre has documented this research-to-deployment gap in detail. It is a significant driver of stalled automation projects, and it falls hardest on the businesses that cannot absorb the cost of a project that doesn’t deliver.

When you prototype in simulation before hardware is involved, the equation changes. Experiments become cheap. Iteration compresses. The decision to buy equipment is made on the basis of something already tested, not a specialist’s estimate. Conversations with integrators suggest a substantial proportion of the projects they are asked to scope never reach deployment – not because the technology doesn’t exist, but because clients couldn’t assess feasibility without incurring the full cost of attempting it.

What would actually move the numbers

For the UK’s SME automation figures to shift, the industry needs to be honest about what hasn’t worked. Grant funding has not been the binding constraint. Training more roboticists is, on its own, too slow and too narrow. Make UK’s research finds that nearly half of British manufacturers cite lack of technical skills as their biggest barrier to adopting new technology. But that skills shortage is a consequence of inaccessible tooling, not an independent cause. You don’t build skills around tools that exclude you.

The more direct lever is widening who can participate in building robotic applications. The production manager who sees the opportunity, the developer who can prototype toward it, the organization that can run a cheap experiment before committing capital. That combination is what unlocks automation for the businesses that have been watching from the sidelines.

We know the robots are ready. But will the people already on the factory floor be given the tools to put them to work? That’s the question.

Nick Thompson
www.olo-robotics.com
Nick Thompson is co-founder and CEO of OLO Robotics, a Sheffield-based startup making robotics development accessible to mainstream software developers and domain experts. Built on ROS2, OLO provides a complete browser-based development platform that combines cloud simulation, AI-assisted coding, visualization and sim-to-real deployment, enabling organizations to build and deploy robotic systems without first becoming ROS2 specialists.