Neural networks help build a sustainable future: curb global warming, preserve nature, develop smart cities and solve social problems. We talk about the most interesting AI solutions for ESG transformation
There are technologies that have changed the path of human development, immeasurably increased human potential and created the world to which we are accustomed today: fire, the wheel, printing, telegraph, computer. In the 21st century, such an iconic technology is artificial intelligence (AI). It not only raises a person’s capabilities to a new level, but also helps to harmonize relationships with the outside world.
Neural networks have become one of the most important tools for achieving sustainable development goals, and today artificial intelligence technologies are making a significant contribution to the well-being of society, business and the planet.
Carbon neutrality and forest conservation: AI to combat climate change
Temperatures on Earth have already risen 1.1ºC above pre-industrial levels, the Intergovernmental Panel on Climate Change (IPCC) says . As a result, extreme natural events are increasingly occurring, causing enormous damage – on average, its economic assessment over the past 20 years has reached $16 million per hour .
Global warming is undermining humanity’s efforts to achieve the UN Sustainable Development Goals (SDGs), experts from the World Meteorological Organization warned in September. It is obvious that the problem requires new approaches and solutions. Nature stated that AI will be able to help achieve 79% of the SDGs thanks to its powerful analytical capabilities. AI provides greater insight into vast amounts of data to develop powerful, simple tools to improve climate resilience and natural resource efficiency.
For example, the online platform Global Forest Watch (GFW) uses AI to analyze satellite imagery to identify forest loss. Thousands of people use GFW every day to prevent illegal logging and forest fires. In addition, GFW helps countries implement national conservation programs.
In addition, AI technologies can synthesize new carbon-neutral materials and optimize production processes to reduce carbon emissions. For example, AI searches for chemical structures that have specific properties, such as the ability to absorb carbon dioxide or retain energy from sunlight.
Thus, AI gives us the answer to one of the global challenges facing humanity: how to maintain the development of production and at the same time reduce greenhouse gas emissions.
Agile Energy Management: AI for the Energy Transition
AI is a critical technology for the transition to clean energy, says Eurelectric study. It is expected that by 2025, 81% of energy companies in the world will implement artificial intelligence. The transition to renewable energy sources, which are characterized by instability, is also increasing interest in these technologies. Therefore, for their effective use, a flexible and decentralized system controlled by neural networks is required.
Experts surveyed by McKinsey also believe that AI will help make the most of solar, wind and hydropower. Algorithms can forecast generation and consumption volumes minute by minute, eliminate electricity losses and ensure the redirection of electricity to where it is needed at the moment. Thus, large green energy projects such as connecting solar and wind power plants in Morocco to the UK power grid via undersea cables could become a reality.
In addition, digital companies themselves, which use a lot of electricity, are reducing their carbon footprint with the help of AI. For example, the AIRI Institute of Artificial Intelligence, with the support of Sber, has developed the open Eco4cast library, which allows reducing CO 2 emissions by up to 90% when performing resource-intensive calculations, including when training neural networks on supercomputers.
And Google is using AI to save up to 40% on data center cooling energy. At first, the cooling system simply used artificial intelligence recommendations. But later, having assessed the effectiveness, all control was completely transferred to AI.
Typhoon monitoring and flood forecasting: AI for climate risk assessment
Artificial intelligence is a revolutionary technology in terms of assessing and analyzing climate risks, says Ecological Processes, because a neural network can provide more accurate forecasts and calculations of the consequences of extreme weather events.
For example, Sber created its own AI model for predicting climate risk events – fires, floods, storms, thawing of permafrost. According to the company, over the past 60 years, fires have become 6.5 times more common, heat and cold waves – ten times, floods – 12 times. The model automates the assessment of financial risks and helps farmers, industrialists and companies from other industries build a strategy for their development, taking into account forecasts. According to Sberbank’s Senior Vice President for ESG Tatyana Zavyalova, such an assessment of financial risks taking into account climate risks will reduce the bank’s losses to 1 billion rubles. in year.
Amur tigers, polar bears and drones in the Arctic: AI and biodiversity
According to the UN, about 1 million species of plants and animals are at risk of extinction, that is, approximately 25% of the entire living world. This highlights the urgent need to improve environmental measures. Artificial intelligence technologies can help here too.
Artificial intelligence opens up fundamentally new opportunities for studying and monitoring the state of flora and fauna. In particular, it will be much faster to process photographic data from drones and camera traps.
One of the solutions in this area is CAPTAIN (Conservation Area Prioritization Through Artificial Intelligence – AI for prioritizing conservation areas). It helps accurately model changes in biodiversity in a specific area due to human-induced factors such as construction. The CAPTAIN service allows you to protect on average 26% more species.
Another direction in this area is tracking rare animals that are in danger of extinction. Portland-based nonprofit Wild Me has developed a neural network model that recognizes zebras and giraffes by their patterns.
In Russia, they plan to use AI to recognize Amur tigers and to count the number of polar bears, walruses, polar gulls and wild reindeer in the Arctic using drones.
Sensors, housing and communal services and smart heat supply: AI for development and smart cities
The UN Economic Commission for Europe reports that cities, as the dominant form of organizing space and people’s lives, play a central role in sustainable development. On the other hand, urbanization also brings with it many problems that slow down the achievement of the SDGs.
For cities to become truly “smart” and sustainable, it is necessary to introduce AI, machine learning, the Internet of things and cloud solutions – the global expert community agrees. “Smart” solutions will allow you to effectively manage resources and reduce the negative impact on the environment.
One of the striking examples of a “smart” city is Oslo. By 2025, they plan to abandon all types of transport except electric ones, digitize all city and municipal services, equip all buildings with sensors, as well as energy and waste management systems based on AI.
In Russia, the Smart City concept is already being implemented in 213 cities with a population of over 100 thousand people. In particular, Rosatom is involved in the development of “smart” cities: the company is implementing AI in the housing and chemical sector to prevent accidents and reduce resource losses.
Sber, in turn, is building a new district in the west of Moscow according to the principles of a “smart” city – SberCity. One of the features of the quarter is a predictive energy management system.
“Smart” energy centers will be located on the upper floors of the buildings. Their special software will monitor the weather and the operation of various devices, such as ventilation systems and heating devices, in real time.
In this way, a comfortable temperature will be created in the houses, and resources will be spent rationally. SberCity also plans to widely use solar panels as facade material – the energy obtained in this way can cover up to 30% of general building needs. The project also provides for the installation of charging stations for electric cars in parking lots. AI will help manage all systems.
Education, medical diagnostics and targeted assistance: AI in the social sphere
The former head of DeepMind and one of the founders of Inflection AI, Mustafa Suleiman , is confident that artificial intelligence will become a tool for achieving social justice and equality thanks to “smart” solutions implemented in medicine, agriculture, as well as for the production of cheaper and more reliable building materials.
AI also has good prospects in the education segment, according to UNESCO. They note that technology makes knowledge more accessible and makes education more individual and of higher quality. AI is also useful as an assistive technology for teaching people with disabilities and health conditions. AI models help them with reading and writing, ensuring an inclusive learning environment.
Neural networks also open up new opportunities in healthcare – “smart” technologies assist surgeons and have learned to select the most effective treatment methods for patients. Various AI models are also being considered for the treatment of depression : they can record changes in the patient’s brain and identify areas that require attention.
And Sber has developed a comprehensive solution based on artificial intelligence that helps diagnose cases of cerebrovascular accidents and provide assistance to patients with stroke as quickly as possible. The service is already used in the Primorsky Territory and is being prepared for launch in the Novosibirsk region.
Finally, some social services are already using AI. For example, social workers in Pennsylvania are using predictive models to proactively provide assistance to families in need. As a result, the number of cases of deprivation of parental rights decreased by 47%.
The Evolution of AI for Sustainable Development
The interaction between artificial intelligence and the surrounding world is not one-way. In other words, it’s not just AI that is having an impact. It may also be the other way around: for example, neural network developers “look” for solutions from nature itself.
According to Semyon Budyonny, managing director, head of the development department for advanced AI technologies of Sberbank, AIRI scientific consultant, sustainable development and artificial intelligence are two large interdisciplinary areas of knowledge aimed at the future. And at the intersection of these two topics, more and more specialists are being cultivated who want to develop technologies to achieve sustainable development goals. Semyon Budyonny emphasizes that now is the time for AI, inspired by both the SDGs and the laws of nature.
For example, the area of swarm intelligence (Swarm Intelligence) for the orchestration of multi-agent systems is developing quite actively. Algorithms are created inspired by the behavior of ants , fireflies and bats . Such models can be used in robotics, route planning and in various industries.
Evolutionary algorithms based on Darwin’s theory imitate the processes of natural selection of living organisms and are suitable for data analysis and process optimization. And artificial neural networks (Artificial Neural Networks), created according to the example of the work of the human brain and the interaction of neurons in it, make it possible to recognize and classify data with enormous speed.
Thus, it is no coincidence that at the world’s largest conference in the field of artificial intelligence, Artificial Intelligence Journey (AIJ), which will be held on November 22–24, a separate track will be devoted to the synergy of artificial intelligence and sustainable development. You can also learn more about how artificial intelligence helps create a sustainable future and creates new opportunities for humans in special podcasts on the AI Journey conference website .
