Artificial intelligence is helping to win the climate change war

Artificial intelligence is helping to win the climate change war

I’m always in favour of better, easier, and more efficient and productive ways of doing things. Which is why I think it is extremely important for us to recognise that emerging artificial intelligence (AI) technologies will almost inevitably become some of our best environmental friends – with two provisos and one personal declaration of interest as I hope to explain!

The first caveat is that we must make sure we really understand what AI is, the full scope of what it can offer, and how to harness its burgeoning potential in almost everything we do.

The second is that we have to accept, for now at least, that AI is also a very heavy energy consumer just as we are going all out to cut energy use – especially fossil-fuel energy – everywhere we can.

Confession or good investment?

I would like to look closely at how AI is reshaping the world around us from the roots up. However, I must admit that I recently took the bold step of investing in a 15kW Powerwall at my own home.

Hopefully, this doesn’t make me unfairly biased! In fact, I will try even harder to be objective.

However, it does help to emphasise the point that while AI is set to change the planet radically, many of its environmental and cost benefits are swiftly becoming routine parts of our everyday lives.

Grid or home-power

My powerwall can store solar power in rechargeable lithium-ion batteries. This gives me the binary-pricing choice of being able to switch out of and into the national grid during demand peaks and lows, and, therefore, the most advantageous prices. AI makes it possible.

I know this is not yet an option for everyone. But scaled up ‘smart’ technologies of this kind in future neighbourhood micro-generation schemes will help to free us from old-fashioned bulk energy sources – gas and oil – that are driving our current energy cost and supply problems.

Also on the glass half full side, one of AI’s other major attractions is that it opens doors to a new generation of environmental services, skilling opportunities, and urgently-needed big picture ‘green’ solutions.

AI roadmap

I thought the best way to explore AI as a key to better environmental solutions is through a couple of definitions and three basic questions.

Later, I look at practical examples of AI in action on global problems – and introduce some Enzygo ( services that could be linked to future AI developments.

Following more examples of AI applications helping the economy, society and environment, I end with a close look at AI’s major downside – its massive carbon footprint.

New definitions

New definitions

The heavy bit! AI systems combine large data sets with intelligent, iterative processing algorithms that can actually learn from patterns in the data they analyse. During each successive iteration, AI tests and measures its own performance and develops additional expertise.

Machine learning, meanwhile, concerns computer systems that learn and adapt with no external instructions using algorithms and statistical models to analyse and take inferences from data patterns.

At the risk of sounding nerdy, I should add that one way of training computers to mimic human reasoning is through neural networks which are a series of algorithms that model the human brain.

Other important digital tools, including AI-driven software, are now part a vast, complex and expanding field which goes far beyond the limitations of this article to explain!

Three key questions

Before moving on to the nitty-gritty of what a ‘smart’ future might feel like, we need to consider some basics points. And one of the first question inevitably asked is what in simple terms is AI?

The broad answer is that ‘intelligent’ computers which use AI to think like humans and perform tasks on their own – and machine learning computer systems which develop their own intelligence – are quickly becoming some of the most powerful tools we have ever invented.

Better design and planning

 The second question is perhaps how can AI actually help us?

 AI and smart systems are set to make a growing difference to our daily lives – not only in future versions of my powerwall, but also in simple tasks such as switching lights off automatically, or sending power stored in resting electric vehicle (EV) batteries back to the grid at peak demand times.

But the sheer volume of data AI provides from around the world in an instant can also help us to understand extremely remote but important changes in the global environment as they happen. This is powerful stuff!

Emergency solutions

However, a third and very practical question is how can AI help us as the current multiple-energy, -environmental and -cost crises gather long-term momentum?

The answer lies in AI’s abilities to gather, analyse, enhance, organise and expand large amounts of ‘big data’, make ‘smart’ decisions on its own, and monitor, model, and manage environmental systems on an unprecedented scale and speed.

For example, to keep modern global temperature rises down to 1.5°C, businesses, policy-makers, and civic leaders need comprehensive information to meet their complex commitments under the 2015 Paris climate change agreement. AI can help to provide this.

Emergency solutions

The Centre for AI and Climate (CAIC) ( recognises this in AI’s potential to ‘push back boundaries’ in climate modelling by processing vast amounts of unstructured data – images, graphs and maps – and open up ‘huge possibilities for understanding the dynamics around sea level rise and ice sheets’.

Practical design and planning aid

We can also use AI data to reduce our use of costly energy. In design and planning applications, AI can help us develop more energy-efficient buildings that account for all inputs, losses and variables.

Renewable energy sources often suffer from intermittency. One answer is to install large-scale batteries in local neighbourhood communities. AI can improve how power storage system operate.

It can also improve green energy distribution by feeding locally-generated solar, wind – and perhaps energy from future versions of my powerwall – into smart grids to smooth out real-time supply and demand fluctuations.

This will be important as we move into a new era of micro-energy generated at home, the office, or a workplace level that can be bought and sold efficiently on a second-by-second basis.

In fact, AI’s potential is so high that a PricewaterhouseCoopers (PWC) study for Microsoft recently estimated could help to cut global greenhouse gas emissions by 4% by 2030 – equivalent to 2.4 Gt CO2e – or the combined projected 2030 annual emissions of Australia, Canada and Japan (!

Greater efficiency

Productivity may benefit too, with faster production cycles, less waste and lower costs. AI could create 38.2 million new jobs across the global economy with an increase in skilled occupations, the PWC study says.

It adds that ‘using AI for environmental applications has the potential to boost global GDP by 3.1% – 4.4%’; environmental applications could contribute up to $5.2 trillion to the global economy in 2030 – a significant leap over business as usual.

Going further still, AI can second-guess human and market behaviour, and predict when vital equipment must be repaired to reduce and even eliminate carbon-emitting failures.

Enzygo (Environmental Consultancy | ENZYGO)

As I have already explained briefly, AI could support environmental service in several ways – including through ongoing developments in advanced mapping and modelling technology.

But the rising availability, quality, immediacy, and breadth of AI data also means more precise inputs to planning practice guidance and submissions to planning authorities.

I see this being increasingly important in flooding assessments, environmental audits, environmental impact assessments (EIA), environmental impact statements, environmental management systems, ground investigation reports, traffic surveys, noise assessments, plus environmental permitting regulations and landscape management advice.

Detailed information on some of our key environmental services where data is important is available at (hydrology), (permitting), (planning), and (landscape).

How AI can help to build a sustainable future

It might be helpful to look next at the full scope of how AI can help society. It clearly isn’t a golden bullet – as yet – because of its carbon footprint. I will look at the downsides separately in a moment.

However, AI’s environmental applications are expanding across the economy – and particularly in precision agriculture; sustainable supply chains; water; energy; and greener transport, including autonomous ‘driverless’ vehicles.

Space-eye view of the world

Using AI and NASA space data, researchers are now identifying in remote regions patterns and land surface changes that have global knock-on effects – such as falling sea and ice cap surface cover.

The potential does not stop there. AI could help halt illegal deforestation, cut water over-extraction, monitor fishing and poaching, reduce air pollution, and improve crop yields. With data from high-resolution radar networks, satellites, and high-altitude balloons, it can close gaps in and improve weather forecasting networks.

AI is even being used to help protect Indian tigers (!

Wildfires and cement

Individual companies are using AI too. Chile’s largest telecommunications company, Entel, is helping to provide a wildfire early warning system using IoT (Internet of Things) sensors mounted in trees as digital ‘noses’ that detect tell-tale airborne particles.

By improving process and quality control, AI is also reported to be helping the Turkish-based cement manufacturing group, OYAK Cimento, to cut its CO2 footprint by circa 7,000 tons annually – the amount absorbed by 320,000 trees.

Food production
Food production

NASA predicts that some crop species will decline by 24% because of climate warming. But food production must rise by 60% to feed two billion more people by 2050. Technology is the key; AI spending in agriculture will grow from $1 billion in 2020 to $4 billion in 2026.

Using AI to analyse local temperature and erosion records, rainfall patterns, soil quality and other parameters, it is now possible to calculate how to maximize yields from small land plots so farmers can use limited resources more accuracy, for example, by not watering crops when rain is expected.

Negative impacts

But the downsides are crucial. AI, algorithms and data processing use masses of energy. The Massachusetts Institute of Technology (MIT) estimates a single large algorithm can result in the release of as much as 284,000 kg of CO2.

Data centres that process and store information linked to sending emails and streaming videos already account for circa 1% of global electricity use, according to the International Energy Agency. By some estimates, computing will account for up to 8% of world’s total power demand by 2030.

The need for rare earth metals in manufacturing AI hardware also has a destructive impact.

Some critics also warn that focussing on tech in the short-term could distract us from better long-term solutions – such as radically changing our lifestyles and taking personal responsibility for our emissions.

Coming to terms with change … again

Throughout our short industrial history, innovations that are unremarkable today were once mysterious and challenging. AI is only likely to be different in its formidable size and potential.

As ever, if more information would be useful, or you would like to discuss any of the issues above in more detail, please contact me directly.

Matt Travis, Company Director, Enzygo Ltd

See the LinkedIn article –

No Comments

Sorry, the comment form is closed at this time.