An electronic device contains many materials — from gold to poison — that shouldn’t end up in a landfill, whether because they’re too valuable or will pollute the environment. Recovering these materials by hand is grueling and dangerous, but the advent of new software and machinery offers hope in the vast sea of e-waste. How can artificial intelligence (AI) recycling and robot technology change waste management?
1. Smart Dumpsters
In 2019, the Americas alone produced 13.1 million metric tons of e-waste. That’s the rough equivalent weight of 19 Empire State Buildings. In a perfect world, people would recycle all their electronic waste. However, significant amounts still end up in dump sites in nations like China, India, and Africa.
It’s much easier to prevent e-waste from going to the landfill than to sort it out afterward. Dumpsters equipped with AI cameras give businesses a rough estimate of the containers’ contents, informing maintenance managers when trash has contaminated recycling bins.
Miami, Florida, has outfitted several city dumpsters with Compology waste-metering sensors. The AI recycling technology will reduce waste collection costs by an estimated 30%–40% and protect the state’s scenic beaches and wetlands.
2. AI Recycling Sorters
E-waste accounts for an estimated 40% of the heavy metals in U.S. landfills. That translates to billions of dollars in valuable metals like gold, copper, and lithium sitting unused at the bottom of a garbage heap. AI recycling technologies make it easier to sort recyclable materials and direct them where necessary.
For example, the TOMRA Autosort Optical Sorter combs through electronics and e-waste on a conveyor belt. It uses ultra-high resolution near-infrared spectroscopy, laser object detection with AI and color detection, and high-resolution metal sensors to sort through waste. A compressor shoots air at objects to remove them from the conveyor belt, ensuring different colored items are in separate bins.
AMP Robotics has developed recycling robots that sort materials much faster than people. The robots’ AI technology processes millions of images, constantly learning to distinguish different types of waste better. The company aims to make recycling more economically attractive by collecting the most valuable, in-demand waste and setting it aside for processing.
3. AI-Based Design
One of the biggest hurdles to recycling e-waste is manufacturers don’t design for it. Most tech businesses create their products with user experience in mind, doing little to address a devi full life cycle.
Companies also make it incredibly hard for users to repair their devices. Simple fixes — like swapping out a dead battery or fixing a broken screen — are difficult or impossible for non-employees. Instruction manuals may be unavailable, and home repairs might void a device’s warranty. Some enterprises even require users to mail in their devices to perform repairs, which is time-consuming and expensive.
Additionally, certain tech manufacturers design their products to fail after a particular period. The batteries might give out, or the software simply stops working. This practice is called planned obsolescence, forcing customers to keep buying new devices year after year.
Devices designed to fail or be challenging to repair are one of the main reasons people only recycle 20% of their e-waste. For example, when a business doesn’t provide repair instructions for a phone, users might damage the device while trying to fix it themselves. It then becomes e-waste. Similarly, products that could last a decade or more end up in landfills after one or two years because of failing batteries or software issues.
Using AI to Prioritize Reuse and Efficiency
Design for disassembly (DfD) makes devices easier to take apart, aiding in recycling. It can turn a throwaway culture into a circular one. AI excels at DfD, finding the best designs for disassembly with ease.
The technology uses a virtual disassembly environment based on two algorithms. One looks at how to assemble the product, while the other analyzes how technicians or machines could take it apart. AI algorithms can explore more design options than a human ever could.
The best designs are modular, meaning swapping out parts when they break is easy. For example, Fairphone makes phones users can repair themselves, right down to upgrading the camera or replacing the battery. DfD products also use less glue or ditch it altogether in favor of large screws, which people can manipulate with a screwdriver.
AI can improve packaging design, leading to lighter, more stackable packaging and less fuel consumption. For example, Apple’s decision to remove the charging block and headphones from its iPhone 12 and later models didn’t just lead to reduced e-waste. It also meant it could ship more boxes on a pallet, saving money, time, and packaging waste in one fell swoop.
4. Other Recycling Robot Technologies
Recovering precious materials from spent devices is much better than mining for new materials. Mining is environmentally, socially, and economically costly, using large quantities of water or acids. Mining can even uncover radioactive waste that requires dangerous and expensive disposal.
In contrast, recycling tends to use less energy than manufacturing new materials — especially when accounting for the machinery used to mine, process, and transport them. So why isn’t everyone doing it?
Recovering the materials from an old phone is only part of the equation — the materials must also be pure enough to use again. Traditional, brute-force recycling methods often degrade recovered metals and plastics by smashing them into pieces. That lowers their value.
Traditional recycling methods can also overlook toxic materials like lead or cadmium, meaning pollutants end up in the ground and cause environmental damage despite going through the recycling process. That is energetically costly, wasteful, and hard on the planet.
The third reason e-waste recycling hasn’t taken off is some manufacturers no longer build devices with screws. Instead, they assemble them with glue, which makes it very hard to take them apart and recover their materials. Strong adhesives can also degrade the value of the materials inside, making it less worthwhile to recover them.
Where These Robots Excel
Recycling robots are game changers in the waste management industry. Apple’s iPhone disassembly robot Daisy can pick apart an iPhone in less than a minute. It uses freezing blasts of air to make the glue inside an iPhone stop working to remove the device’s battery. Then, it removes the phone’s tiny screws and separates the remaining components.
A smartphone contains several elements that are costly to mine and process. Daisy can remove aluminum, plastic, cobalt, rare earth metals, copper, steel, glass, tantalum, gold, tin, lithium, tungsten, paper, and zinc from a phone. This recycling robot can disassemble 200 iPhones an hour, and — unlike a human worker — it’s always up for the job.
Recycling robots save companies on mining, labor, and manufacturing costs. They also improve corporations’ environmental reputation and make it easier to procure necessary materials — no more waiting for shipments from other countries. Recycling ensures manufacturers always have enough material to work with. With many mineral supplies dwindling, that’s good news for an enterprise’s bottom line.
The best solution for recycling e-waste would be an AI recycling robot that could work on any device. Someday, a jack-of-all-trades robot could be a reality. However, manufacturers build electronics with different layouts and materials, meaning sorting robots must be device-specific.
Pushing for a Circular Economy
Reducing the volume of electronics people generate in the first place is one of the most important ways to address the e-waste crisis. Design for disassembly, the right to repair, comprehensive disassembly instructions, and banning planned obsolescence will go a long way.
Ultimately, however, the most vital step in e-waste management will be changing consumer spending habits. Until people decide not to buy the latest iPhone model every year, the sheer volume of electronic garbage will be too much to handle — even for a robot.
