Climate change is real. More real than we might want to accept.
The average global temperature is rising by 3% every year now; 20% of the species are on the verge of extinction; the global sea level has risen by 8 inches in the last century – the scientific evidence of climate change is unmistakable but there’s hope. While the perilous effects of climate change have snowballed through the years, the good news is that technologies like machine learning can help solve it.
The AIML approach to this phenomenon is powered by data. Hence, it’s safe to say that the reports, predictions, and preventive planning are more likely to work.
Let us look at some ways in which Artificial Intelligence & Machine Learning is keeping climate change in check.Effective Climate Change Predictions-
Accurate climate predictions are crucial to prevent any more damage. Climate informatics, an intersection of data science and climate science covers a wide range of climate information reports – predictive reports, probability reconstruction using historical data, down-scaling or up-scaling of weather models, and much more.
Artificial intelligence and machine learning tools draw inferences from a copious amount of data, making the reports near-accurate. Climate modeling has benefitted from AIML tools and has grown in scale through the years. Whether it is the oceans, land, or atmosphere, the climate reports generated by AIML tools have over time proved to be dependable sources of information. However, climate scientists suspect that these reports forecast work only for short-terms and there’s room for improvement.
Needless to say, these predictions help nations and organizations to prepare well for natural disasters, understand the socio-economic impacts of climate change better, make more efficient climate policies, and recover from the damages. The more accurate the reports, the better our chances of saving the planet.
Renewable source of energy is not only a sustainable model but also a safe option for the planet. Wind and solar energy, are in fact the best options for generating electricity. Not only are they cheap, but also easy to harness. However, there are several road blockers in the path of harnessing renewable energy. Engineers often face issues with predicting weather changes, calculating demand-supply ratio, surplus planning, and more. Machine learning algorithms process humungous amounts of data to gain insights on energy requirements and ways of meeting those energy demands. Whether it is real-time weather conditions, data from the solar panels, pollution metrics and more, ML algorithms can turn these data into useful information. AI can also monitor and operate power plants according to optimal weather conditions.
Machine learning tools control and maintain networks of supply, storage, demand and more by using algorithms based on bandit theory (which decides the choice to maximize reward). Through this method, we can maximize power generation or minimize emissions.
Transportation networks can also benefit from machine learning tools. ML algorithms used by cab services can be effectively used for business cargo and cut down on fuel consumption and emission.
Physics-based ML models are now increasingly used to forecast new material behavior. These models further help the conservation of energy by predicting properties and compatibility between multiple sets of materials.Showing the effects of extreme weather-
Not every part of the world has faced the effects of climate change and a rise in temperatures, just yet. But some homeowners have already faced the severity. Researchers from Montreal Institute for Learning Algorithms (MILA), Microsoft, and Consistent AI Labs have started to make use of GAN’s (generative adversarial network). This is a type of AI that helps stimulate what homes would look like if they were damaged by the rise of sea levels and storms.
MILA Researchers have met with NGO’s and Montreal City Officials to discuss the use of this AI tool. They are eager to help people understand what their neighborhoods would look like if they were damaged by climate change. The tool is not yet fully functional. It would need to collect more data, but they plan on making an app where people can upload pictures of floods and fires. This data can then be further used by the algorithm.
Victor Schmidt, a Ph.D. Candidate at MILA said that their goal is not to convince people that climate change is real but to get people who believe in it to do more.
Monitoring carbon emissions with the use of satellite imagery, carbon tracker uses gathered data to convince the financial industry that coal plants are not profitable. It is an independent think-tank that is working towards the UN’s goal of stopping new coal plants from being built by the year 2020.
AI can be used to automate the analysis of images of power plants and get regular updates on emissions. New ways to measure a plant’s impact by tracking numbers of nearby infrastructure and electricity use can also be introduced with the help of AI. Carbon tracker will track about 4000-5000 plants in order to gather more data than it currently has, thus improving it’s functioning.
Using water without really paying attention to how much we may be wasting has become part of our daily routine. But in today’s world, we need to become more aware of our consumption levels. An average American household uses about 320 gallons of water per day, and 30% of that goes into maintaining landscapes.
Smart Irrigation systems that use AI technology, can help save up to 8,800 gallons of water per home, per year. And about 120 gallons across America.
Skydrop is one such product, it is run on a wireless internet connection. It stays up to date with local weather and collects data so that it knows when it should not water your lawn. For example, during a downpour. It is similar to Nest and gathers data based on the customer’s preference rather than just being set on a timer. It also uses technology which can help tell us the moisture level in the soil.
All of us use heating and cooling systems in our homes and offices on a day to day basis without knowing the impact. They account for almost half the average residential energy use and demand a lot of energy usage as well. Nest has come up with a smart thermostat known as Nest’s Learning Thermostat. By using this, we’ll be able to save money and the planet too.
Smart thermostats basically are able to automatically adjust your indoor temperature settings based on the data of external humidity. The device uses a Wi-Fi connection in order to be able to check what the outside weather is like. The AI tools are also able to track user behavior and make manual adjustments based on previous use and learn our preferences and become better at adapting the temperature to its user’s needs over a period of time. Using this saves at least 10% of energy on heating and roughly 15% on cooling, per year.
Artificial Intelligence can improve energy efficiency by incorporating data from smart meters and the IoT (Internet of things) by enabling devices to send and receive data to help forecast data.
AI systems can simulate potential zoning laws, flood plains, to help in being prepared in case of a natural disaster. In China, IBM’s Green Horizon project is using AI systems that are able to forecast air pollution, track pollution sources and produce potential strategies to deal with it.
AI is able to simulate climate and explore strategies to test how to ease heat waves. If a city wants to plant new trees, this AI model will be able to determine the best place to plant these trees in order to get optimal tree cover and reduce heat.
- Google: In a statement, Google explained how it’s efficiency has improved and how value-wind farms are improving by 20% with the use of Artificial Intelligence to schedule energy deliveries based on predictability models. As a result, they are less reliable on fossil fuels.
- SilverTerra: Powered by the funds and technology provided by Microsoft, SilverTerra is able to use AI and Satellite imagery to predict the size and health of forest trees. This can be used to ensure that trees remain healthy and continue to grow well.
- DHL and IBM: 23% of the global greenhouse gas emissions are caused due to transportation. DHL and IBM have teamed up to use AI to be able to improve DHL’s global logistics operations. This can be very beneficial, as, between 1970 and 2004, there was a 120% increase in emissions of the transport and logistics sector.
Check out how Google’s DeepMind is using AI to tackle climate change.
There is No Planet B!
It’s reassuring to know that Artificial Intelligence and Machine Learning has help us avert and even reverse the effects of climate change. Researchers all around the world are closely following AIML trends with the hope that changes can be implemented sooner than later.
These are just a few ways through which Artificial Intelligence can be used to fight climate change. The future of our planet depends on how we are able to make use of the advancement in technology to make a difference. You can up skill with Great Learning’s PGP in Artificial Intelligence and Machine Learning.