Reporting 2020

Can AI do climate protection?

How can data help protect the climate – and what green opportunities do new technologies offer? We talked with Simone Kaiser, Deputy Head of the Center for Responsible Research and Innovation (CeRRI) of the Fraunhofer Institute for Industrial Engineering IAO, about artificial intelligence and the wrongly perceived contradiction between ecology and economy.

Interview with Simone Kaiser

Ms. Kaiser, how can AI projects be applied to support sustainability?

Artificial intelligence has great potential for climate and environmental protection. It can be an important instrument for achieving our sustainability goals. The opportunity it presents is that it can process large data volumes, rapidly identify complex interrelationships and learn from them. That has three advantages: We can make processes more complex and simultaneously more precise, optimize processes regarding sustainability indicators and reach automated decisions. We can benefit for example in mobility, using AI to calculate climate-friendly routes. AI can also help avoid traffic jams or pool people or goods together in vehicles.

What are the obstacles for AI in the context of sustainability?

AI is fundamentally a neutral technical tool. It is not inherent in the tool whether it’s eco-friendly or instead promotes climate-destroying developments. The job of researchers, developers and users is to apply AI in such a way that maximizes its green and climate-protecting effect. There are also concrete risks we need to keep in mind. It takes a lot of power to run computer centers and to train the algorithms. The hard­ware has to be constantly upgraded in ever shorter cycles. We’re all familiar with that from our smartphones. Obviously, all this has an environmental impact. Some people point out that digitalization is not yet systematically used for the protection of nature or the climate, but more for conventional growth. That’s something we all need to work on.

“Artificial intelligence has great potential for climate and environmental protection. It can be an important instrument for achieving our sustain­ability goals.”

Simone Kaiser, Deputy Head of the Center for Responsible Research and Innovation (CeRRI), of the Fraunhofer IAO

Do you think it would make sense to examine AI projects from a sustainability point of view from the start?

Definitely. I don’t see a contradiction between economic and ecological benefits. Climate change is also a big danger to the economy. The market is increasingly rewarding sustainable business models, so there’s a large economic potential. We can’t consider one aspect without the other.

This calls for interdisciplinary work, bringing together people with sustainability expertise and AI expertise. How can we do that successfully?

That’s one of our research focuses and a vital step toward transferring sustainability-oriented AI projects into practice. At the moment, the two viewpoints are not often considered together. People who work in sustainability relatively rarely use AI and many developers and startups don’t know how their AI could impact on ecological challenges. A tech expert who develops an AI for image recognition doesn’t necessarily realize that it could be used to spot illegal clearances in rainforests on satellite images. We need to create networks and platforms that enable cooperation and give people more insight into and understanding of each others’ disciplines. I firmly believe innovation requires looking at a problem from different perspectives. This variety of view­points can be collected and combined via managed processes so that everybody benefits from all kinds of expertise and joint solutions can emerge.

What about ethical aspects? For example, how should we deal with the fact that AI can make some activities obsolete?

Above all, it’s important that this discussion doesn’t just take place among experts. Companies should deal openly with their employees’ worries that they might lose their jobs or that their jobs will change so much as to be unrecognizable compared to today. To help in this process, we develop visions of the future: What could the industry look like in ten years’ time, and how might new technologies be used then? We invite people to formulate their wishes for this future. That opens up new prospects for shaping the future. This is how we can get people to set off on a path even though we don’t yet know where exactly it will lead. Sure, it takes courage, but it also offers great opportunities.