When I’m talking to organisations about digital manufacturing, how to reduce energy usage is a really hot topic – particularly quick ways to reduce energy usage.
According to the 2019 World Energy Outlook by the International Energy Agency (IEA), “a sharp pick-up in efficiency improvements is the single most important element that brings the world towards the Sustainable Development Scenario”.
We know that for a growing number of organisations, reducing energy usage is about more than cutting costs. They’re recognising the importance of building sustainability into their operations, and value the way this enhances their reputation. Moving to digital manufacturing and using AI and machine learning to drive Industry 4.0 energy efficiencies also showcases an organisation’s commitment to innovation, something that can help them to stand out from the pack when it comes to bidding for contracts.
So, what’s the easiest way for companies to use AI to reduce energy usage and potentially improve their carbon footprint?
AI adoption depends on the ‘people factor’
The first thing I’d say is that successfully implementing AI in an organisation is more about how you manage the people side of things than anything else. AI has huge potential in digital manufacturing, but if people don’t trust it, and feel that it’s being done ‘to’ them, it’s never going to be that beneficial.
I don’t believe in imposing AI on people or enterprises. Instead, we support people to use a standardised algorithm to build their own model using their own data – or even to find ways to build on work they’ve already started. We find that if people can see how the logic works for themselves and are fully engaged in building the model, they’re much more likely to deploy and trust it. We think this is the only way to get to the point where the AI is doing more than just giving guidance and is allowed to run automatically. I’ve written more about the importance of building trust in an earlier blog post.
AI that coaches you to reduce energy usage and carbon production
Working with our partners, we’ve developed algorithms that quickly adapt to your environment to help you reduce your energy use and your carbon footprint. Typically, we’re seeing energy savings of 6-12% within 1-2 months.
The AI works like a sports coach, looking at your past operations, modelling the best outcomes, and giving recommendations to help you achieve your efficiency ‘sweet spot’. With our technological partners, we start by identifying all the variables that have an impact on your energy consumption, such as your machinery settings. Then we plot an Energy Efficiency Index (EEI) to look at past production runs, comparing them to see which were the most energy efficient and why. We also compare how sites and equipment perform to look for best practice.
From this baseline, the AI can provide predictions and recommendations for specific production runs, suggesting optimal operational parameters and indicating how much you could save in costs and carbon consumption. Every time the AI’s used, it learns, continually refining its recommendations. And when you’ve learnt to trust it, the AI can make adjustments automatically, within the safety limits you’ve defined, becoming a closed loop solution to optimise energy and carbon emissions.
It’s dynamic, too. The model is aligned to existing energy standards and will constantly review your performance, recommending adjustments to make sure you keep achieving the highest efficiency ratings.
Breaking free from ‘pilot purgatory’
I’m aware that a lot of organisations have fallen victim to what McKinsey calls “pilot purgatory”, where they’ve got significant activity underway, but have yet to see any meaningful bottom-line benefits. Our approach is all about breaking free from this scaling issue, so we’ve focused on working with our partners to create AI that can be easily adapted to different environments.
For example, with our partners, we’ve helped a steel mill save $2.8m a year, stopping the creation of 41,600 tonnes of carbon emissions – from just a single furnace. And it’s easy to feed different data into the algorithm to use it in other industries. For instance, if you swap out the steel mill data inputs with wind speed, tidal conditions and propeller speed for a ship, you’ve got a model to reduce the use of marine diesel. We’ve done exactly that with one of our maritime customers and they’re now seeing a 10% saving. We’re working on other use cases, too, ranging from heavy industry to buildings and data centres.
Leading by example
As a business committed to achieving net zero by 2045, our road to reducing emissions from our operations includes switching to renewables, decarbonising our estate and transitioning to a low carbon fleet. Over two-thirds of our end-to-end carbon emissions come from our supply chain, so we’re also working with our suppliers to reduce their carbon emissions. We’re proud to say we’re on track, having recently hit our 100% renewable electricity target. And, since 2009 / 10, our energy efficiency programme has unlocked savings of over £343m.
We’re eager to help more organisations with energy and carbon innovation so, if you’d like to find out more about running a trial in your business, please get in touch with your account manager.