5 lessons on Industrial AI from Hannover Messe
What an F1 simulation car, a robot dog, and three honest conversations at the CONTACT Software stand taught us about getting AI right in UK manufacturing.
Industrial AI is being sold as a feature. The teams that succeed treat it as an operating-model decision.
Walk any major manufacturing trade show in 2026 and AI is everywhere. Bolted on top of an MES module here. Promised as a roadmap item there. The pitch is always the same: "switch this on and your data starts working harder."
In practice, the UK manufacturers we work with are finding it is rarely that simple. AI is only as useful as the data, the workflow and the workforce that surround it. The teams getting real value treat AI as a deliberate change to how the business operates, not as a feature flag on an existing system.
"AI is no longer the cherry on top of the cake. It is the engine that makes the digital platform worthwhile in the first place."
Why AI is suddenly on every UK manufacturer's roadmap
Cost pressures, skills shortages, sustainability requirements, and the steady rise of global competitors that are already further down the digital road. AI is being pitched as the way to keep up with all of it at once.
The risk is buying a capability before you understand the problem it solves. The reward, for businesses that get this right, is a step-change in decision speed and operational visibility that compounds over years.
What the F1 simulator and Conny the robot dog showed us
The stand. At Hannover Messe, the CONTACT Software stand revolved entirely around Industrial AI and Digital Transformation. The eye-catchers were a fully networked F1 mini simulation car and Conny, a robot dog.
The F1 car. Fully connected. Every lap streamed historical and simulation data back to the driver, who could see exactly where they were losing time. The fun part was learning that the gap between "instinct" and "data" is not as wide as you think when the data is good.
Conny. The robot dog inspected the car after every test drive and suggested material and design improvements. A small, contained example, but the principle is the one that matters: AI is most useful when it has the context to give a specific recommendation, not a generic one.
The conversations. Off-stand, I spent time with Sergio Viera Velasco and Frank Vandenberg comparing notes on what makes Industrial AI projects land. The five points below are the ones that struck both of us as genuinely useful, regardless of which platform you are working with.
AI is most powerful when it understands the context of your business.
The single biggest lesson from the conversations at Hannover Messe: AI in a manufacturing setting is only as useful as the surrounding context it can reason over. Product structures, part variants, engineering change history, machine state, operator decisions. The richer the context, the more useful the output. The thinner the context, the more generic and therefore useless the suggestion.
That has a practical implication. Before you switch on AI, you have to get your data house in reasonable order. Not perfect. Reasonable. The teams that try to bolt AI onto a fragmented data estate are the ones that end up disappointed.
5 lessons on Industrial AI, distilled
The five points that came out of our conversations at the CONTACT Software stand. Each one is something we are now actively applying with UK manufacturing clients.
Treat AI as an enabler, not the objective. Your strategy should be built around business outcomes, data readiness and organisational capability, not algorithms or vendor demos.Embed AI in your workflow. Do not bolt it on. AI is at its most powerful when it understands the context: product structures, variants, requirements, engineering changes. A model that sits next to your workflow is a curiosity. A model that is woven into it is a capability.Look at your organisation with fresh eyes. Question your data management, your team capabilities, your change-management plan. What data do you have, where does it live, and who owns it? Get clear answers before you spend money on AI.Plan in years, not weeks. A serious Industrial AI programme is multi-year. Split it into manageable steps that each deliver a tangible business benefit. Plan early wins to keep momentum, and accept the plan will change as you learn.Get specialist help. You do not know what you do not know. Specialist help can range from a one-hour benchmarking conversation to full project design and implementation. Either way it pays back.Bring AI to your connected products
Smart connected products with closed-loop engineering data. Apply the lessons from Hannover to your manufacturing operation.