Artificial intelligence – the answer to protecting the environment?

Artificial intelligence – the answer to protecting the environment?
29 Jun 2021

The 20 warmest years since records began occurred in the last 22 years. Global heating is accelerating and is becoming more noticeable. It is an urgent problem – if not the most urgent problem – of our time, and many changes are needed to pave the way for a sustainable society. The demands of the environment and the economy are repeatedly weighed against each other and an ideal balance seems to be utopian thinking. So how can we bridge the gap between profitable and sustainable trade?

One frequently discussed option is artificial intelligence (AI). This has enormous potential for positive effects at both a commercial and an environmental level. We have summarised some of the possibilities for you here.

The temperature difference from 1884 to 2020 (Source: NASA/GISS)

Different AI applications

Artificial intelligence – which immediately makes you think of science fiction – is already well-established in many areas. Contrary to popular belief, AI is not a new phenomenon and has been applied in practical situations for a long time. Essentially, AI means machine learning – intelligent learning procedures enable processes to be continuously and independently optimised and adapted. Machines and software systems with AI are able to work on particular problems or tasks without instructions from a human. Advances in digitalisation result in increasing numbers of potential applications.

Prognoses

As long as no unforeseen events occur, AI is able to make more or less reliable predictions, for example, about future consumption of products. This enables companies to optimise materials purchasing using data gathered from experience.

Chatbot

Chatbots are based on the principle of identifying individual keywords and automatically executing actions associated with them. This helps them to give customers the feeling that they are having a natural conversation.

Data analysis

AI can significantly improve and simplify the evaluation and analysis of unstructured data. Content can be compared, classified and checked in a focused way.

Maintenance

By using automated remote error detection, AI can detect problems early on and resolve them. This makes equipment maintenance significantly easier. Consequently, it can extend the equipment’s service life and save avoidable costs.

Using AI for sustainability and the environment

Natural disasters such as drought, floods or storms throw nature out of balance. How can at-risk ecosystems be recognised in advance and protected?

AI methods open up new possibilities for natural disasters. AI is capable of finding patterns in highly complex data and recognising potential solutions. For example, large quantities of data from Earth monitoring systems can be evaluated and predictions made as to the effects of natural disasters. Decision-makers can access this information very rapidly. Questions about how healthy an ecosystem is, for example, can be answered quickly and then action can be taken to make improvements.

AI methods can also make an important contribution to sustainable development in the waste disposal and recycling industry. Rubbish is automatically separated and sorted via algorithms. A robot gripper arm equipped with sensors can precisely differentiate and separate a variety of materials. The technology promises huge potential, especially for recycling plastics. The process is expected to offer an improvement in the recycling rate in the future.

Digitalisation is the cause of an increasing number of problems, as well as ever-growing mountains of electronic waste. Currently, only 1% of critical components and rare earths are recycled properly. Using artificial intelligence as a tool would enable development of more efficient recycling processes. What’s more, based on how easy it is to recycle a product, recommendations could be made as to how the product could be designed to be more sustainable in the future – and to use fewer resources.

The data-intensive computing power necessary for AI uses up a lot of energy. But, at the same time, AI can also help save energy. Throughout the entire manufacturing process, AI can conserve resources and even avoid using them in certain places, reducing the burden on the environment enormously.

The downside of artificial intelligence

Despite its many advantages, AI is not a panacea for protecting the environment. As mentioned above, AI applications require large amounts of computing power, so they themselves need a lot of energy and resources. In “The Green AI”, the Allen Institute for AI in Seattle investigated the energy consumption of artificial intelligence. The results showed that between 2012 and 2018, the computing power doubled every few months. Towards the end, it grew by 300,000 times – and needed correspondingly increased amounts of energy. According to a study by the University of Massachusetts, the learning process emits five times as much CO2 as a car during its entire lifetime. However, most of the energy is consumed during an artificial intelligence’s initial phase. Once it has been trained and is working, it needs a lot less energy.

Conclusion

Of course, artificial intelligence won’t lead to an increase in sustainability automatically, but it does offer enormous potential and possibilities. If AI is applied as a tool in the right places, it could certainly have a positive effect on protecting the environment and our finite resources.


Lisa Neulichedl