Toxic waste is becoming a huge problem that AI can solve

Four hundred million tons of toxic waste are produced every year.

Imagine that for a second – that’s it 70 Great Pyramids of Giza combined. It’s not something most of us consider when ordering a white chocolate mocha latte in a plastic cup at Starbucks, but waste management is a huge problem for an entire industry charged with preventing environmental damage.

Failure to address that problem results in things like rising methane emissions, which are responsible for 25% of climate change. Aside from climate change, proper management of waste management has a direct impact on the safety of the general public at an immediate level.

Household Hazardous Waste (HHW), such as flammable or reactive products such as paint, cleaners, oil, batteries and other chemical products, poses a serious risk if not disposed of properly. Despite this, between 20 and 80% of all household waste generated is dumped into open spaces, bodies of water or sewers. What remains is burned or buried, according to the UN.

There simply aren’t enough people to effectively separate recyclables from general waste, or toxic waste from other waste. If you can catch the gist here, this is where artificial intelligence (AI) can help – and not at the expense of anyone’s job.

The power of image annotation

AI enables a process known as image annotation that allows a machine to effectively cluster parts of an image that belong to the same object class. In plain language, this allows waste management facilities to use machines to identify all the recyclable material being processed. They can then separate it to be used again. The system can identify all toxic waste more effectively than a human.

Annotators use small lines connected to vertices to trace the shape of objects in digital images. In addition, annotated video helps AI algorithms understand motion and label each frame of video training footage. Correctly identifying the shape of streets and roads is useful for computer vision models trying to figure out if the garbage has been dumped incorrectly. Lane annotation adds lines to images and frames to represent the structure of linear objects such as roads and railways.

AI-powered camera systems can also help cities combat illegal waste dumping. Using object recognition capabilities, street cameras can identify garbage objects cluttering sidewalks and roads and alert city authorities to send workers to clean them up. Street-level cameras may also be able to capture license plates used by dumpers, which could lead to fines and act as a deterrent.

The fact is that government regulation is not seriously considered if the problem it is trying to solve seems unfeasible. And until AI emerged to address the waste management problem, regulation on a global scale was simply inadequate.

The challenge of regulating the runoff of HHW

The framework for dealing with hazardous human waste (HWW) in the US was established by the Resource Conservation and Recovery Act of 1976 (RCRA). Those standards, of course, only apply to the US market. There is no global standardized definition of HHW, meaning that an item may be hazardous waste in one country but not in another. To make matters worse, most cities cannot afford to provide enough manpower to enforce proper removal of HHW.

In many countries, hazardous waste is not subject to legislation unless it is separated from other household waste. So it is difficult to create the right legislation to address this. Therefore, hazardous and non-hazardous waste are often mixed together and not thoroughly sorted before disposal.

From the waste that is released during, for example, care activities 85% is generally non-hazardous waste. But 15% is considered material that may be infectious, toxic or radioactive, according to the World Health Organization. Workers in sorting plants are then modestly exposed to toxic waste, leading to major health problems over time.

The issue of HHW is not to be taken lightly. If left unchecked, there are serious consequences that can be fatal. Not only sorting plants (and those who employ them) are affected, anyone in a city using drinking water can feel the reverberation.

Especially when you consider how few jobs it would actually cost – if any – the benefits of AI in this industry outweigh the downside.

Arie Zilberman is CEO and co-founder of Key Makr.