Can modern technology contribute to environmental protection and have a positive impact on changes taking place in the world? Can artificial intelligence save the world? How is machine learning used currently in the service of ecology, and how can IT companies contribute to a more sustainable development?
The study of endangered species
Artificial intelligence can not only improve the reduction of the negative impact of human activity on nature but also track the effects of species protection. AI is used, for example, for the non-invasive study of animal behavioural patterns, such as migration, mating, and eating habits. The Footprint Identification Technique (FIT) allows scientists to collect data on endangered species of animals, providing data from photos of animal footprints at the individual, age, and sex level – without any interference related to trapping and tagging animals.
Air quality monitoring
Software such as City Air Quality Management helps cities with implementing more effective plans faster and improving air quality. It allows one to predict the expected pollution several days in advance and suggests solutions that can reduce the occurrence of smog. For example, it suggests limiting the use of trucks or other vehicles with combustion engines in certain areas with smog problems.
The software is extremely precise. For three days ahead, its accuracy is as high as 90%, and for five days ahead – 80%. Based on the level of pollution, City Air Quality Management can distinguish between working days and weekends, and even in its calculations, it takes into account events that take place cyclically, such as sports events or other mass events. The solution is used by Stuttgart and Nuremberg.
The world population is growing rapidly. Food demand is expected to increase by 70% by 2050. Therefore, agriculture should become more productive and precise to ensure high yields and quality with the least possible use of harmful chemicals. To meet these challenges, many farmers use a variety of intelligent devices including sensors and cameras that allows them to process information and monitor numerous parameters in real-time, such as irrigation, plant nutrition or the occurrence of diseases and pests.
Using hyperspectral cameras mounted on drones, researchers can measure the state of chlorophyll levels and the level of the greening of crops. The more chlorophyll, the stronger the yield is – this information can be used in the fight against fungi. Moreover, the collected data is also used to calculate the size of the harvest, which in practice means that the producer can avoid overproduction.
Blue River Technology has developed a technology that utilises cameras and information processing by artificial intelligence algorithms to accurately monitor each plant in large-scale plantations. Due to the data collected, an AI-controlled machine can selectively spray herbicides directly onto weeds without contaminating the plant. Information provided by Blue River suggests that their solution eliminates 80% of the volume of chemicals that are normally sprayed on plant crops, which in turn is expected to reduce Hebrides spending by 90%.
Forecasting weather changes and natural disasters
Since 1970, the number of natural disasters has quadrupled. This resulted in the death of over 3.3 million people and enormous material damage. It is estimated that by 2030, 60% of the world’s population will live in cities, and 1.4 billion people will live in places most at risk from natural disasters and extreme weather events. Climate science does not remain passive and is developing towards using data, machine learning to be able to predict weather changes in the world.
Machine learning ensures the maximum accuracy in weather forecasting, especially in the field of extreme phenomena such as hurricanes and cyclones. The aforementioned enables the creation of many models and simulations that are extremely useful. For example, with data showing rising sea levels, we can predict larger and more violent storms in certain areas.
Ebb and flow monitoring
Wave dynamics is a key part of our global earth system. Ebb and flow not only determine the life of fauna and flora but also have great economic importance. However, current climate models do not directly analyse waves due to cost and complexity. The National Oceanographic Center of Southampton uses grants to try to predict wave states in the North Atlantic through the use of deep learning mechanisms.
Using cost-effective computational methods allows you to build wave models and mimic interactions with the ocean and atmosphere. Collected, processed and analyzed data requires computing power that goes beyond ordinary spreadsheets. Besides, deep learning allows for more effective and faster inference, pattern building and finally learning of information processing machines.
Improving transport efficiency
More efficient engines, better aerodynamics, hybrid or electric engines and weight reduction of the vehicle are just a few options on the market that significantly improve transport efficiency. These proposals require the application of engineering techniques using machine learning and artificial intelligence. We are talking about the construction of an internal combustion engine, improving the operation of electric hybrid vehicles, modelling turbines for aerodynamic vehicles or additive production – this allows the production of lighter parts for vehicles and ultimately reduces energy consumption.
However, the greatest hopes for increasing the efficiency of transport can be seen in autonomous vehicles. Machine learning is essential in creating vehicles of this type of unmanned cars because it offers a large-scaled implementation of road tracking, obstacle detection, energy reduction and eco-driving to such vehicles.
Energy without a carbon footprint
Today, electricity systems account for more than a quarter of greenhouse gas emissions. Machine learning can help reduce their emissions. First of all, it facilitates forecasting electricity demand, as well as supporting flexible supply management so that generators do not generate excess electricity. Another application, especially important during the transition from standard coal to low-carbon energy, is the better integration of alternative energy sources with the power grid.
Thanks to automatic detection, AI can also reduce the leakage of methane (a gas more dangerous to the climate than carbon dioxide) in pipelines or switching stations. Machine learning will also find application in the so-called transfer learning, which means the possibility of using data from intelligent power grids in systems that do not have advanced tools to monitor their operation. Such information transfer will allow them to optimise their operation despite the lack of data.
IT companies help protect natural resources
Large-scale commercial and industrial systems such as data centres consume huge amounts of energy, and while much has been done to reduce their demand, much remains to be done as the demand for computing power grows globally. Reducing energy consumption has been an important goal for big cloud service providers. They invest in building super-efficient servers, developed more efficient ways to receive energy from devices in data centres, and invested in renewable energy, with the goal of 100% coverage of energy demand from renewable sources. For instance, Google now has about 3.5 times more computing power than 5 years ago, for the same amount of energy consumption. However, the company continues to develop new enhancements to further reduce demand.
Of course, going green goes beyond data centers. Future Processing is an example proving that even a smaller IT organisation can have an impact. The eco-friendly office, which is situated on a brownfield, offers special parking space for bikes with basing mending equipment as well as parking spaces for electric cars. Reducing greenhouse gases can be seen in garbage segregation and points for collecting caps and waste electronic equipment. Besides, in the company’s HQ, you can find a canteen with veggie meals. What is more, posters on saving water and reduction of used paper towels can be seen everywhere.
Call to action
At first glance, warming the climate by a few degrees may seem insignificant to us, but in fact, this relatively small increase in global temperature may lead to the melting of the polar ice, and thus – raising the level of the oceans, increasing the frequency of hurricane waves and heat waves, reducing the biodiversity of our planet and, as a consequence, the destruction of ecosystem structures. Thus, we are in desperate need to act now. In this context, IT-based solutions can play a significant role in this race against the clock – both in micro as well as macro scale.
Do you have an idea about how to make your company green? Join our expert analysts and developers to bring your idea to life.
<!–[if lte IE 8]>
Source: Future Processing