AI and Big Data: Transforming Sustainability in Manufacturing – ReGen Write Up

It’s a big topic and one that cannot be ignored. As AI systems continue to evolve, literally bringing new functionality and greater efficiencies to business each day, our next panel explores how it is transforming sustainability in manufacturing.

Chairing the panel, Rachel Swann, Director of Rachel Swann Ltd was joined by Lawrence Dudley, Co-Founder of Parallax, Dan Graf, CEO of Earthchain, Ahmad Mahmoud Kobeiter Abiad, Sustainability Specialist at New Wave Biotech and Paul Howell, Founding Director of Arquella Ltd.

Once introductions are made, the panel are asked whether sustainability, when delivered by new technologies, is a financial decision?

Paul – we use millimetre wave technology to track residents’ movements. This could be the speed from a chair to the bathroom. What this allows us to do is to predict the likelihood of a fall.

Measuring the response of this is imperative as care is funded outside of the home. The overall budget is recognised when we can reduce the number of people going into hospital by reducing accidents.

Right at the point of usage there might not be financial incentive, but at the ecosystem there is. My challenge is how you sustain that. There needs to be a joined-up process.

Dan – there is almost always a financial system. Sustainability is a beautiful side effect of being efficient.

Lawrence – with energy being quite expensive, it means there is a whole world of stuff you can do that has a return. You can do things that are financially viable, it’s just reporting that.

Ahmad – when it comes to fundraising, people have to show off the environmental credentials of the product. Whatever that product is, you have to show the numbers. It is all about the external fund driver.

Dan – you need to include the figures when you complete a tender too. It’s the sustainability measures. If you don’t have a carbon reduction plan, then you will fail. Showing AI in practice, for us it was able to turn 5,000 data points around in 48 hours. It did tens of hundreds of calculations in a second.

Lawrence – I have seen many poor examples of carbon reduction plans. The assumption is that the grid will get greener and your carbon reduction plan is done. I think we need to invest in tangible things that are going to deliver a real return.

We can say that we put solar panels on the roof to be sustainable, but the real reason is that it will reduce costs.

The panel then moved on to explain how they ensure that people use AI to make the manufacturing process more efficient.

Paul – manufacturing is incredibly simple in principle terms. It’s about efficiency in the supply chain. I worry that financially driven AI decisions are short sighted. If we become long-sighted and look at critical bottle necks in the supply chain and conversation we will become world class manufacturers. If we don’t, we won’t.

Lawrence – it’s difficult to set out long term plans in manufacturing. It comes down to whether you can do it better, faster and cheaper. Its hard for an individual to set out where they will be in 20 years when they don’t know where they will be in three.

Paul – we forget that just because we work in manufacturing, it doesn’t mean we have to manufacture. This is where start-ups are winning time and again. They have a blank piece of paper and they can set themselves up to be the most efficient they can be.

AI will be revolutionary over the next ten years. As it becomes focused on niche solutions that are for that specific business it will have an enormous impact.

Lawrence – problems and optimisations in manufacturing are very specific. We are some way off having niche solutions for that. Some problems are so unique and complicated you need creativity to solve them.

Dan – my personal view is that when people enter a role in sustainability, they want to make the world a better place. The point is people enter careers in this sector to save the planet. They don’t want to be in a bunker somewhere reviewing data.

With desk jobs you can get a number wrong and you’ve messed up a report and you are looking at data that isn’t even real. I want to see people expanding on strategy and really implementing it.

As AI is power heavy and hungry, the panel were asked if we are not exacerbating the problem with sustainability by automating everything?

Paul – there are several concerns for me. We are sticking data into the hands of the big four. In the future, we will be looking at deep seek and those that use less electronic requirement.

We are talking about systems that will use 5% of the digital resource. We have fragmented intelligence from individual knowledge resources, so these tools are brilliant at some things but terrible at others. AI will do the same.

Ahmad – in general we are not going to turn our backs on AI as it is such a useful tool. We just need to make it less resource intensive. We are focusing less on the energy use and more on the cost savings. There are some businesses that simply wouldn’t be viable without these tools.

Lawrence – there is a cliff edge. Most providers of AI tech are operating at huge costs. We don’t know the true cost until we have the true pricing. When it needs to generate cash in return we will get the answers to those questions.

Raising the point of ROI, the panel turned to the fact that we don’t usually get the actual cost of something. We don’t take into account all its different parts. The question was, do we need to look at how we determine investment against ROI?

Dan – speaking very candidly, ROI gets a deal across the line. No one will ever work it out or go back to check.

Paul – I think our start-up world and the whole 20 years has lost its focus on what business really is. Ai is a reality that we haven’t worked out yet. We haven’t worked out the cost or the pay off. We’ve seen they are emerging, and they are there to use, so we feel we have to implement them. But we don’t know what is down the line. We will go through the same as we have with social media – AI will be great for humanity and terrible for humanity.

Dan – again, it’s a cliff edge. It might be the great reaping. After all, 90% of exposure to AI is people using the tools to produce memes. It is a result of AI eating itself. If it is precious and the resource is precious, we need to think carefully about how it is used. Ultimately, the cost will be shared amongst people.

Lawrence – they will make it more efficient, it will make money, and it will all be fine. People will pay for it like any other software. We need to stop thinking about it as anything other than software. Are businesses going to pay, of course they are.

Paul – I don’t think it is like any other software, I think it’s the first time in the software journey that huge swathes of companies are giving free insight into their data or their operations for free, with little to no thought.

Lawrence – if you use AI in enterprise, you will have licensing to stop customer data being used.

Paul – I don’t think data is problematic, I think the way in which we come to the answer, which is our intellectual property, is what we need to be mindful of. The way we learn to do something, which is the intellectual property of business is being commoditised or centralised to a certain number of organisations. The data is safe, it’s the methodology.

Dan – it was the same in the industrial revolution. We took away the artisanal skills which was one head knowledge.

Paul – that became 10k mills rather than 4 organisations around the globe. Which is why I believe the next critical stage of IP and competitive advantage will be to retain the data and the methodology of how the data is created.

What that does for the larger organisations is an interesting question. When we get new model generation, then we start to commoditise the building of the models.

Dan – if you know what questions to ask, it can do incredibly impressive things.

In the last part of the session, the talk turns to uploading data into the hands of people we don’t know. The questions posed was, what could go wrong and should we be worried?

Dan – we all buy smartphones and others buy Alexas and that is far more insidious. They are listening to you. These things are listening to you. I don’t think it’s any more sinister, it’s just more of the same thing.

Paul – I rest my case. If Alexa is listening, then every one of the deep models is listening in a deeper and more insidious way. There will be another wave of individuals that will break out and gain transactions with IP protection and security will be incredibly important.

We’ve been involved in 3D printing since the early 2000’s, it was NASA technology that was commercialised. What’s happened is that it has become commoditised because others are introducing systems at a much more competitive rate. That transition will happen with AI as well.