In Germany, Artificial Intelligence Is Making Wind Turbines More Bird-Friendly

In Germany, the rate of construction of new wind turbines has slowed dramatically - which could have fatal consequences for the energy transition. One of the reasons that new turbines face opposition is that the giant blades endanger near-flying birds. BirdVision is tackling the problem with the help of AI.

Author Sarah-Indra Jungblut:

Translation Sarah-Indra Jungblut, 02.16.20

In Germany, the rate of construction of new wind turbines has slowed dramatically – which could have fatal consequences for the energy transition. One of the reasons that new turbines face opposition is that the giant blades endanger near-flying birds. BirdVision is tackling the problem with the help of AI.

With climate change continuing apace, we need to reduce our fossil fuel consumption and switch over to renewable energies as fast as we can. But in the biggest country in Europe, the growth of wind energy – an incredibly efficient source of energy that the country will rely on to reach its renewable energy targets – has slowed to stagnation point in the past few years.

There are numerous reasons why – including changes in government subsidies, the time it takes to issue permits for new wind turbines (it used to take just 10 months, but now can take over two years), and also unclear targets for planners with regard to land allocation and legal disputes. In addition to all that, all around the country, civil society groups are forming and filing lawsuits against planned sites – complaining about the increased noise level and also the visual impact that wind turbines have on the surrounding landscape. New regulations take these factors into account, meaning that no new turbines can be built within a certain radius of where people are living. The result of this is that there are hardly any spaces that are suitable for new wind turbines (all the more absurd when you consider all the things that are allowed to be built closer to residential buildings than a wind turbine: including motorways, coal-fired power stations, airports, landfills…).

Another point of criticism is the suggestion that wind turbines threaten wildlife, which is repeatedly cited by both citizens’ initiatives and nature conservation organisations. In recent years, all this has led to the refusal of permits or to considerable restrictions on the construction of new turbines.

“This is where we want to take action, providing the industry with a camera system that detects bird species flying too close to the wind turbines, and pausing them before they do any harm. The aim is to avoid endangering any wildlife at all,”

Benjamin Friedle, one of the three managing directors of BirdVision, explained to RESET. Since the beginning of 2018, the partners have been developing the “BirdVision” camera system together with the Bürgerwindpark Hohenlohe, phil-vision GmbH (a company that specialises in image processing) and Dr Michal Lewandowski, a specialist in neural networks and tracking. The system is designed to detect and track bird species that are potentially at risk of harm from wind turbines and protect them by switching off the turbines as soon as they approach.

BirdVision – Protecting Birds With Artificial Intelligence

BirdVision is a specially-developed camera system: high-performance outdoor industrial cameras are mounted at the foot of the wind turbine and connected to an image processing server. With the help of a neural deep learning network, birds that are potentially at risk of harm from wind turbines are detected via intelligent tracking. If one of the endangered birds approaches the turbine, the system registers it and switches off the blades as soon as it gets near. The servers are located directly in the tower of the wind turbine, they drawing their power directly from the turbine itself.

In 2019 BirdVision was tested on eight wind turbines and so far, the team is satisfied with the results. The results are currently being compiled by external evaluators and will be published in a report, but according to Friedle, they’ve been able to achieve very high detection rates so far and the company plans to start small-scale production by the end of 2020. There is also allegedly a very high demand from industry: “particularly among project developers who are applying for new permits and finding it hard to get them approved and operators of existing wind turbines who are currently using very cost-intensive avoidance measures,” says Friedle.

Short pauses, not long shutdowns

Wind turbines in France and Spain have already been fitted with similar camera systems – but those don’t have the added advantage of AI. “We believe that when it comes to detection rate, false trigger rate and reliability, they simply can’t compete with systems that use artificial intelligence” says Friedle.

When you imagine the huge rotor blades found on industrial wind turbines it seems like it must be a huge technical feat to get them to stop. And surely bringing them to a halt has an economic disadvantage too, because when they’re not moving, no electricity can be produced? The managing director of BirdVision disagrees: “Technically, we’ve been carrying out an identical shutdown for over five years when we turn off the blades to protect local bat populations. Depending on the weather, we have between 300 and 600 shutdowns in southern Germany during the summer months. Of course, no electricity is produced during the few minutes that each shutdown lasts. But that demand-oriented technical shutdown is already being offset by very cost-intensive blanket shutdowns and avoidance measures in countless wind farms. We would like to replace those with this technology.”

BirdVision is successfully bringing together wildlife protection and environmentally-friendly energy production. While at the same time, the team’s solution hopes to make more land usable for more wind energy – land that is currently not available because of the dangers posed to local wildlife. “It’s the job of federal and state legislators to work with the wind energy industry and nature conservation organisations to develop more practical and simpler solutions, of course” says Benjamin Friedle. Following BirdVision’s lead, let’s hope they will.

This is a translation of an original article that first appeared on RESET’s German-language site.


This article is part of the RESET Special Feature “Artificial Intelligence – Can Computing Power Save Our Planet?”

The RESET Special Feature on AI is part of a project funded by the Deutschen Bundesstiftung Umwelt (German Federal Environmental Foundation DBU). As part of this project, over a period of two years we will be developing four RESET Special Features on the topic of “Opportunities and Potentials of Digitalisation for Sustainable Development”.

You can find more information here.

Masakhane: Using AI to Bring African Languages Into the Global Conversation

Researchers from across the continent are collaborating on an open source AI project to develop machine translation for African languages - facilitating communication, increasing accessibility and opening doors to the world’s youngest continent to play a stronger role in shaping the digital world.

Carbon Intensity Forecast: How Green Will Tomorrow’s Electricity Be?

A emissions-tracking tool, developed in Britain, is using machine learning to work out if the next 96 hours of electricity has been sustainably produced - offering a big boost to both energy-consumer transparency and the transition to renewables.

Winnow: Harnessing the Power of AI to Tackle Food Waste in the Catering Industry

From farm to plate, unnecessary waste is a huge problem throughout the food supply chain. London-based company Winnow has come up with an AI-integrated waste disposal system to help restaurants dramatically reduce what gets thrown away.

Bractlet: Creating Digital Clones of Buildings to Help Save Energy

Like every person, every building is unique. So how can you easily, affordably and reliably reduce the energy consumption of something so complex? An American start up delivers a digital solution.

Sustainability and Digitalisation in the Spotlight: Receives Project Funding from the DBU!

The official confirmation arrived just a few weeks ago. We're excited to announce that over the next two years, the German Federal Environmental Foundation (DBU) will provide RESET with both technical and financial support to develop an in-depth feature series that takes a critical look at future digital sustainability challenges. And kick-off is next week!

German Artificial Intelligence Researchers Hope AI Can Spot Land Damage Before It’s Too Late

Satellites armed with artificial intelligence could act as a high-tech early warning system for environmental damage.