Artificial intelligence (AI) is transforming almost every industry – from medicine, finance, education and customer service to manufacturing and transport. It is also a true revolutionary when it comes to climate change, environmental protection and sustainability. Here are 6 eco-related areas where its potential has a bright future.
The Bin-e intelligent waste bin that segregates rubbish on its own is the beginning of a change when it comes to waste management using AI. The bin uses machine learning to identify, categorise and sort waste as soon as it is thrown away. It recognises and sorts rubbish into 4 fractions and tells you when the bins need emptying (I mentioned it HERE).
But until all waste bins are smart, waste must be sorted at a waste management facility. And here we are back to day-to-day reality. Is that really the case? No! There are already machines powered by artificial intelligence that can sort 160 materials per minute (the human result is 30 to 40 recyclables per minute). Moreover, they can work around the clock.
Such a solution has been developed by Greyparrot, among others, which has used AI-based vision software to increase transparency and automate recycling. The way it works is that items are scanned with cameras and analysed by deep learning algorithms. See for yourself.
Let’s go further! Every year, billions of dollars are wasted on properly disposing of or recycling used parts from electronic devices. Apple has started using recycled or low-impact materials in the manufacture of its products. As a result, these products can be reused.
What about food waste? Expired food products are also a huge amount of waste thrown in the bin every day. An Israeli startup thought about this and came up with an algorithm for dynamic pricing of food products. It is called Wasteless.
The price of a product is closely related to its expiry date – if the end date is approaching, the product is automatically discounted. This means that the price decreases day by day and this encourages people to buy cheaper food products, which prevents them from becoming expired and ending up in the bin. Wasteless can also be integrated into a shop’s inventory management system. As its developers argue: by using Wasteless, a retailer can reduce food waste by 39%, while increasing revenue by 110% and maintaining a positive net margin.
WATER and OCEANS
Artificial intelligence can be used to protect the oceans from illegal fishing, for example. Satellite data and vessel traffic data are used in a machine learning algorithm called Global Fishing Watch. It allows illegal fishing to be identified.
In turn, the Ocean Data Alliance uses AI to track water pollution, ocean mining and coral fading. With near real-time data, decision makers and authorities are able to respond to problems much faster.
Meanwhile, the Nature Conservancy is working with Microsoft to use artificial intelligence to map the wealth of the oceans. Assessing the economic value of ocean ecosystems will enable better decisions on protection and possible areas for action, such as fishing or mining. Maps and models have already been created for Micronesia, the Caribbean, Florida and the project is expanding to Australia, Haiti and Jamaica.
How else can we use artificial intelligence in this area? With it we can predict the spread of invasive species, track marine debris, monitor ocean currents and measure pollution levels.
And what about the water outside the oceans? WINT is artificial intelligence for detecting and stopping water leaks in water supplies. The system does this by using pattern matching. The AI in WINT is constantly learning and adapting to the network. One company that uses WINT has confirmed that AI has reduced water leaks by 24%. This is always a good beginning.
FORESTS and ANIMALS
Artificial intelligence is used, for example, to conduct non-invasive studies of animal behaviour patterns such as migration, mating and feeding habits. An interesting tool for collecting data on endangered animal species, for example, is Footprint Identification Technology (FIT). This is software developed by WildTrack, which collects animal track data at the level of individual, age, sex – without any interference from trapping and tagging.
Global Forest Watch, on the other hand, was developed to protect biodiversity in the Amazon. The app displays alerts that show where deforestation, illegal gold mining, and logging are occurring from protected areas. As estimated, such early warning systems have the potential to save 32 million hectares of forest worldwide by 2030! I wrote about deforestation and the danger it poses to the planet in an article HERE
Other similar Earth apps using artificial intelligence include iNaturalist and eBirds. They collect data from various experts on the species they encounter, which helps to track their populations, ecosystems and migration patterns.
Another project, called PAWS (Protection Assistant for Wildlife) from the University of Southern California, uses machine learning to predict where poaching might occur in the future. The algorithm currently analyses past ranger patrols and poaching behaviour based on crime data.
A Long Live the Kings in Washington state is trying to restore declining steelhead and salmon populations. Thanks to a grant from Microsoft, the data collected will help support habitat protection for these fish and restoration efforts.
ENERGY and RES
Did you know that Google’s own artificial intelligence, called DeepMind, has helped the IT giant reduce its energy consumption? Google used machine learning to predict when energy in data centres was most in demand. The system analysed and predicted when users were most likely to watch YouTube videos, for example, and then optimised the cooling needed. As a result, Google reduced energy consumption by 40% in the data centres making them more energy efficient and reducing overall greenhouse gas emissions.
AI along with IoT and Big Data can help maximise profits. Wind companies, among others, are using it to drive the propeller of each turbine to generate more electricity, taking into account real-time weather (wind direction) and operational (propeller settings) data. In contrast, the Department of Energy’s SLAC National Accelerator Laboratory, operated by Stanford University, wants to use machine learning and artificial intelligence to identify vulnerabilities in power grids from solar PV farms to protect them from outages and help restore power faster when they do occur. Researchers will analyse data from renewables, energy storage and satellite imagery. The aim is to develop a solution that can automatically manage renewable energy without disruption and help the system recover quickly from outages with little human involvement.
A startup in Berlin has developed an application called Plantix based on deep learning that identifies potential plant defects and nutrient deficiencies in the soil. The software’s algorithms analyse photos of leaves, whose appearance correlates with specific soil defects, plant pests and diseases.
Two US companies Where and FarmShots are already using machine learning algorithms in conjunction with satellites to predict weather, analyse crop sustainability and assess farms for disease and pests. So is Microsoft and the US Department of Agriculture’s FarmBeats pilot project, which feeds data from sensors, drones, satellites and tractors into cloud-based artificial intelligence models. These provide a detailed picture of soil quality and moisture levels in fields located on a 2,800-acre research farm in Maryland. Sensors measure temperature, moisture, acidity and water levels in the soil. The weather station records air temperature, precipitation and wind speed, while the sensor-equipped tractor in turn records elevation, biomass and crop ‘greening’, an indicator of plant health. If the project is successful, farmers will be able to view the data generated by FarmBeats in real time, which could help them improve their farming methods. If all goes to plan, the system will then be tested on more than 200 farms across the country.
Do such projects make sense? Absolutely. A similar solution in India has helped farmers achieve 30% higher peanut yields per hectare by providing information on land preparation, fertilisation and choosing sowing dates.
What’s more, by combining AI and automation in agriculture, it is possible to monitor crop health in real time, detect livestock diseases faster, programme field irrigation and herd feeding, and thus reduce the use of water, fertiliser and pesticides.
CO2 EMISSIONS and AIR
In China, an IBM project called Green Horizon uses an artificial intelligence system that can forecast air pollution, track its sources and develop potential coping strategies. For example, it can determine whether it would be more effective to reduce the number of drivers or close certain power plants to reduce pollution in a particular area.
The report “How can artificial intelligence enable a sustainable future?” (How AI can enable a sustainable future?), which is a joint effort between Microsoft and PwC, estimates that the use of artificial intelligence for environmental applications could increase global GDP by 3.1 – 4.4% while reducing global greenhouse gas emissions by approximately 1.5 – 4.0% by 2030. As the authors predict in the report: the productivity benefits of AI applications in four key sectors, such as agriculture, energy, transportation and water management, could generate overall economic growth, yielding a potential return of $3.6 – $5.2 trillion. At the same time, widespread AI applications can accelerate the transition to a low-carbon world, reducing global greenhouse gas emissions by 0.9 – 2.4 gigatonnes, equivalent to the annual emissions of Australia, Canada and Japan combined in 2030.
I am very glad to see that “machine learning” and AI are developing dynamically to better protect dying animal species, to help segregate waste more accurately, to manage energy, water or air pollution. This gives hope that in this age of technological revolution we are also thinking about the planet.
If you are interested in the topic of artificial intelligence in the context of the environment, I recommend reading the Microsoft and PwC report I mentioned earlier [to be read HERE] and the World Economic Forum report “Using Artificial Intelligence for the Earth”. “Harnessing Artificial Intelligence for the Earth” (Harnessing Artificial Intelligence for the Earth) [to download HERE ].
For a broader view of the application of AI in business, it is worth reviewing the study “AI. Challenges and Consequences” by Infuture Institute and Natalia Hatalska [HERE].
I am very much interested in your insights on this topic.
Please write, comment, ask!