AI Being Used In Monitoring Illegal Waste Disposal

Artificial intelligence is slowly being implemented in satellite images to observe farms and monitor their activity. The artificial intelligence system is being tested in the United States to track down farms that may be illegally polluting the waterways or dumping waste where it should not be disposed.

In the US, there are many facilities known as concentrated animal feeding operations (CAFOs) consist of around fourty percent of the country’s livestock. These intensive farms very often accommodate as many as 2500 pigs or 125,000 chickens per facility, generating approximately 335 million tonnes of waste per year. The problem with this mass amount of waste is where and how to dispose of it. Although some is disposed of in a way that does not harm the environment, some farms illegally throw waste into the river, polluting it and causing harm to marine life and villages that happen to use the river to fetch resources.

Manure is largely responsible for a great proportion of this waste, which often makes its way into waterways untreated. In the United States, the Clean Water Act states that, anyone who wants to dump waste into a waterway requires a federal permit, however the Environmental Protection Agency (EPA) estimates that nearly 60 per cent of CAFOs don’t have one.

Many farms in the US dispose of waste illegally.

Many farms in the US dispose of waste illegally.

Currently there are no federal government system that track the number and size of these farms, so environmental groups have to compile their own data by manually scanning satellite images which is highly inefficient.

Daniel Ho and Cassandra Handan Nader at Stanford University trained a neural network AI to scan available satellite images for CAFOs. The AI managed to identify certain distinctive features, such as rectangular barns or outdoor manure pits. The algorithm found 15 per cent more farms than were previously known, which gives an increase of 589 farms. The efficiency of the AI is far beyond that of human, it managed to scan and report CAFO’s in two days as opposed to an average of six weeks by a human.

Similar techniques are slowly being implemented in agricultural land across Europe. These algorithms are monitoring the health of vineyards in southern Italy, and observing farmers in Lithuania and Estonia who receive government subsidies for keeping their land in good condition.

The AI technology used Estonia detects whether farmers are mowing their fields as required of them. This has made it easier for inspectors as it helps reduce the need for physical visits from inspectors. The implementation of an AI system like this is estimated to save £500,000 every year in manual inspection.

The future is looking better with AI being implemented in instances like this where farmhouses contribute a great amount to the removal of waste illegally. Not only does this technology save costs but from an environmental perspective, it helps preserve and keep our environment cleaner and free from potentially hazardous waste. As this technology becomes more advanced and with the help of the government, a push for this AI to be implemented in many more farmhouses can be done faster.

AI scans satellite images for farms that dispose of waste illegally.

AI scans satellite images for farms that dispose of waste illegally.

Zacharia Sharif