The Industrial AI investment opportunity

After decades in the high-growth and innovation space, I’ve seen more than a few hype cycles come and go. I lived through the dot-com boom and the market correction that followed, and I’ve watched (or sometimes led) wave after wave of new technologies that promise seismic industry transformation. Experience teaches perspective. It reminds you that while the headlines and social marketing chase noise, the real, lasting value tends to emerge quietly in the companies that solve hard, meaningful problems with discipline and purpose.

I believe we’re in another of those moments now with AI. Generative or “clerical” AI dominates both attention and investment, but the tide is already turning, with significant questions about ROI at the enterprise level.

By contrast, the quieter but far more consequential revolution is happening in Industrial AI - the application of machine learning to engineering, manufacturing, energy, and other complex 'multi-physics' systems. That’s where genuine productivity gains, measurable ROI, and durable advantage lie.

When you unleash the potential of machine learning at mission-critical engineering, the ROI is tangible rather than theoretical. One of the projects the Monumo team is working on has driven a 17% reduction in a customer’s bill of materials compared to the original electric motor design. That equates to potential savings up to €70 per motor. Extrapolate that saving to the average motor production run of a European OEM, and you are talking tens of millions in savings on just one component. That kind of economic value starkly contrasts with the 'nice-to-have-but-hard-to-pay-for' clerical productivity gains that generative AI buyers experience.

Back in the early 1800s, Brunel's bold ambition and unwavering belief that he could solve challenging problems helped define a British culture of invention and investment. That culture remains strong today with world-class talent across the country, deep-tech investors who understand patient capital requirements and defensibility, and globally relevant sectors such as automotive, energy, defence, aerospace, and med-tech that are ripe for tools to make them faster and more efficient.

Industrial AI doesn't require tens of billions to train frontier models, making it suitable for the UK’s scale and capital mix, and you might say 'understated Britishness'. The media would now have you believe that we are on the brink of a bursting AI bubble. While it is true that some businesses, particularly in the generative AI space, are overvalued, this doesn't mean the category is mispriced overall. As we saw in surgical robotics, a few platforms will deliver outsized returns while many followers won’t. For VCs and their LPs, the filter is simple - can the customer rationalise the business decision to buy? Is there a clear go-to-market strategy? Adoption and measurable ROI will beat hype and vanity metrics every time. No doubt, some big-name VCs, most likely in San Francisco, will get out quickly and see history-defining returns. My view, however, is that for the long tail of investors to see big payoffs, they need to divert attention to Industrial AI and recognise that the sector's compounding returns take time and therefore patience to play a slightly longer, less hypey, but ultimately more impactful game.