SwissCognitive Guest Blogger: Neil Kitching, geographer and energy specialist from Scotland

Artificial Intelligence is the use of technology to sense their environment, process, problem solve, learn and take action from the analysis of large volumes of data. All AI uses technology built from materials and minerals mined from the ground whilst all AI uses energy to function. More on this later.

AI to monitor the environment

The most obvious environmental use of AI is to deploy sensors to gather information to monitor the environment. For example, sensors in the sewage network can be combined with data from rainfall gauges and weather forecasts (also from AI) to predict sewer overflows in advance. Action can then be taken to divert water flows if possible or to warn residents of an imminent flood risk. The same sensors can monitor the condition of the pipes, detect corrosion and warn of tiny leaks before they grow and burst to enable planned and cost effective maintenance and replacement of buried assets. AI can also use satellite data to analyse and identify changes in land-use such as to spot illegal logging or to monitor changes in soil carbon. This can be combined with payments to ‘offset carbon’ – paying farmers to manage land in a way that protects and enhances soil carbon.

The role of AI in transportation

AI is being used in many transport applications. Intelligent traffic lights, monitoring congestion and optimising routes for passengers and freight will reduce air pollution and carbon emissions. And AI is the backbone of proposed driverless technology. Proponents dream of a future with self-driving electric cars and suggest that this will solve many environmental problems. If we shared such cars and used them constantly then we would need fewer cars. There would be less congestion, less time wasted searching for a vacant car park space and less land taken up by parking spaces. The software could select the most fuel-efficient route and speed. Computer controlled cars with automatic brakes could drive close together reducing wind resistance. But access to self-driving cars is likely to result in an increase in total distance travelled and may increase congestion. For the first time, everyone would have access to a car, people could commute much further and work or sleep during the time spent travelling. In any case, people like to own status goods like cars and might still choose to own a self-driving private car rather than hire one when they need it. Like all technology the outcome depends on what people choose to do – behavioural issues that people who develop technology do not always think through.

Another view of AI in terms of Sustainability

The problem with claims that AI will reduce environmental impact and carbon emissions is that the knock-on consequences are rarely considered. Technology allows us to do more, often more efficiently at a lower price so inevitably demand and consumption increase. Using the clever algorithms used by companies like Amazon, we can now order goods cheaply from anywhere in the world delivered to our door. More stuff, more emissions.

Technology hardware and batteries are made from materials like plastic, silicon, cobalt and lithium. Mining is energy intensive and environmentally destructive and is concentrated in a small number of countries. In the Atacama Desert in Chile, miners inject precious scarce water into underground pools to force the saline water to the surface to concentrate the lithium through evaporation. Some of this water is produced from energy intensive desalination then pumped up mountains to where it is needed. Chile has more than half all known reserves of lithium, whilst other metals such as coltan come from countries prone to conflict, corruption and use of child labour such as the Congo. Without improved efficiency and recycling, our society is in danger of becoming as reliant on these countries as we currently are on oil from the Middle East.

To operate AI requires power, mainly electricity. All connected technology requires electricity to operate the device; to transfer data between the device and servers; and to operate data centres where data is stored. Globally the electricity used to operate this network emits more carbon than aviation. And the volume of data being collected, processed and stored is growing fast. Energy efficiency is improving too, but struggles to keep up with this exponential growth. Obviously the use of renewable electricity is preferable to electricity powered by fossil fuels, but all electricity generation has some form of adverse environmental impact.

New technology, which I cannot claim to understand, has the potential to massively increase electricity consumption. Crypto-currencies require vast amounts of electricity to produce. And any claims made that blockchain will open up ‘sustainable supply chains’ need to be considered alongside the energy required to operate blockchain.

Is AI a force for Sustainable Good or not?

To conclude, AI has many uses that will improve efficiency and reduce carbon emissions. Yet, we also need to work to optimise the design, architecture, storage and energy efficiency of the technology behind AI.

About the Author

Neil Kitching is a geographer and energy specialist from Scotland. He has written his first book, Carbon Choices on the common-sense solutions to our climate and nature crises. He works for a public sector agency promoting the opportunities for business to benefit from low carbon heating and water technologies.

Der Beitrag AI – A force for Sustainable Good? erschien zuerst auf SwissCognitive | AI Ventures, Advisory & Research.

copyright by grist.org

The United Nations has called on the world to protect 30 percent of the planet from human activity to help protect ecosystems and slow down climate change. But conservation areas are often vulnerable to illegal logging, poaching, mining, and other activities that threaten biodiversity. How can land managers detect these kinds of human impacts on protected ecosystems? Scientists are applying machine learning to identify human influence on the environment by literally listening to the environment — that is, by monitoring forest “soundscapes.”

Every ecosystem has its own distinctive collection of sounds that change with the season and even the time of day. According to Bryan Pijanowski, soundscape ecologist and director of Purdue University’s Center for Global Soundscapes, “Sounds are part of the ecosystem, and they are signatures of that ecosystem.” The unique sound environment of an ecosystem is known as a soundscape, the aggregate of all the sounds — biological, geophysical, and anthropogenic — that make up a place.

Sound has long been used by soundscape ecologists to assess biodiversity and other metrics of ecosystem health. Pijanowski has his own, informal rule of thumb: “If I can tap my foot to a soundscape, I know it’s fairly healthy,” he says, because it means “the rhythmic animals — the frogs and the insects, the base of the food chain — are there.”

New research published in the Proceedings of the National Academy of Sciences applies tools from machine learning to these soundscapes to get a better picture of ecosystem health and human activity. The researchers built algorithms that taught themselves to predict habitat quality in different environments across the world, ranging from rainforests in Borneo and the Republic of Congo to temperate forests in New York, based only on sound data.

Detecting human activity that impacts ecosystem health, like illegal logging and poaching, has long been a challenge for land managers and scientists, often requiring expensive and time-consuming surveys in which specialists manually identify species. But this new method requires only basic audio equipment that allows for remote monitoring of the soundscape, which can be done in real time, and a machine learning algorithm that listens for sounds that aren’t typical in a forest environment. “Say that there’s weird things going on or illegal activity, like guns being shot, or chainsaws from illegal logging,” explained Sarab Sethi, a mathematician at Imperial College London and the lead author of the new paper. “We work under the assumption that illegal activity contains a lot of anomalous sounds that are different from whatever usual sounds are in the ecosystem.”

How does the computer identify strange sounds? The key is unsupervised machine learning, meaning machine learning that doesn’t require human input to “train” the model on pre-identified data. “The way that we measure similarities and differences in sound is really the technical advance from our work,” Sethi told Grist. This new method uses a neural network to compare the “fingerprints” of sounds — not only their frequencies, but the structure of how their frequencies change over time — to one another other. “Once we’ve got a fingerprint, like a bird calling — a bird calling is more similar to a different species of bird calling, in this fingerprint, than it is to, say, a gunshot,” says Sethi. The neural network learns which sounds are typical of a healthy forest environment, and which ones are out of the ordinary.

The unsupervised technique requires less work from humans to identify sound; it’s also more robust than so-called supervised machine learning. Unsupervised, the algorithm detects anomalous sounds on its own, without requiring a fallible human researcher to teach it what gunshots and chainsaws sound like. “If you use a supervised approach, your whole approach succeeds or fails based on how good your training data is, so how well labeled that data is,” said Sethi. “You don’t have that sort of reliance in unsupervised methods.”[…]

Read more: grist.org

Der Beitrag Scientists are using machine learning to listen to nature (literally) erschien zuerst auf SwissCognitive | AI Ventures, Advisory & Research.