The human existential crisis: AI has the potential to tackle climate change

AI algorithms can help predict the right mixture of crops to plant to regenerate soil health and reduce fertilizer use

The human existential crisis: AI has the potential to tackle climate change - CIO&Leader

Climate change is the most important crisis the planet is facing today. Millions of people from all over the world took to the streets recently demanding urgent governmental action to help control the ongoing catastrophe and reverse the negative impact of climate change. We will need to marshal all our resources, including Artificial Intelligence to save our planet from peril. Some of the foremost minds in machine learning and artificial intelligence recently published a study where they outlined 13 crucial areas where machine learning can be used to mitigate the adverse effects of climate change. The recommendations they made were divided into three major categories – high leverage solutions, where machine learning can make a noticeable impact, long term solutions that will take at least a couple of decades to pay off, and finally, high risk pursuits, where the technology is either not mature enough or we don’t know enough to effectively predict the consequences. Policy measures by the government should focus on implementing some of the high leverage solutions at scale to reduce our carbon footprint. Let us take a look at some of them that can make a significant difference:

  • Improve predictions on electricity needs – To deploy renewable energy in a more efficient way, utilities must be better able to predict energy requirements, both in real-time and over a longer time period. There are plenty of algorithms that can already predict energy demands, and we can refine them further by considering climate patterns, local weather, and household energy usage. By making the algorithms more understandable, we can greatly assist utility operators interpret outputs and use the results to schedule when to switch over to renewable energy in the power grids.
  • Discover more energy efficient materials – Scientists are constantly looking for ways to harvest, store and utilize energy in a more efficient manner. However, the process of discovering or synthesizing new materials is slow and not very precise. Machine learning can help speed up the process of designing novel chemical compounds with energy efficient properties. For instance, it can help us design solar fuels that can capture and harness the Sun’s energy more efficiently, identify materials that absorb carbon dioxide better and design structural materials that take a lot less carbon to manufacture. The day is not far when the latter would replace cement and steel, which will be a change for the better, as their production accounts for nearly 10% of greenhouse gas emissions.
  • Design energy efficient buildings – Intelligently designed control systems can drastically reduce the energy consumption of buildings. They can be programmed to take building occupancy, weather forecasts, and other miscellaneous environmental conditions so residents can adjust the cooling, heating, lighting and ventilation needs indoors. Smart building can also directly communicate with the electricity grid and reduce power usage if there is paucity in low-carbon power.
  • Help scale up precision agriculture – Monoculture currently dominates agricultural practice. It refers to the practice of growing one crop in a vast area. Adopting this approach is easy for farmers as they can deploy tractors and automated tools to manage the upkeep of their fields. However, it reduces the nutrient content of soil and hampers productivity over time. Consequently, farmers have to heavily rely on nitrogenous fertilizers that produce nitrous oxide, which is one of the most potent greenhouse gases. By deploying robots that use machine learning based software, farmers can manage a mixture of crops more efficiently, and AI algorithms can help predict the right mixture of crops to plant to regenerate soil health and reduce fertilizer use.
  • Improve the tracking of deforestation activity – Deforestation is a major contributor to the emission of greenhouse gases. However, the prevention of deforestation is a laborious on-ground manual process. Satellite imagery combined with computer vision technology can help us scan for the loss of green cover more efficiently and on a much larger scale. Sensors, if planted strategically on the ground, can be used in combination with algorithms to detect axe and chainsaw sounds. This will greatly help law enforcement officers detect and put an end to illegal deforestation activity.

For people in the field of machine learning, this is an exciting time to use their skills for the greater good. Identify opportunities in your community and outside where you can use your skills in any way to help build a low carbon economy, and contribute to the global fight against climate change. 

The author is Founder, Thatware LLP

Add new comment