Introducing X's moonshot for circularity

Humans generate tons of waste.

To be more precise, by 2050, studies suggest we’ll produce 3.8 billion tons of waste per year—a 73 percent increase from 2020. And as trash continues to accumulate on our planet at an accelerating rate, managing it becomes even more unsustainable.

Mismanagement of waste is a climate problem that exacerbates so many intersecting environmental issues. Landfills emit potent greenhouse gases, persistent plastics clog our oceans and other vital ecosystems, and microplastics are a growing societal concern to human and ecosystem health. The more we throw away, the more we need to produce new materials from scratch–usually using fossil fuels.

But what if the world’s waste could actually become our most valuable resource? Our team of scientists, researchers, and engineers at X have been hard at work devising radical approaches to our planet’s ballooning waste management problem.

As X’s moonshot for the circular economy, we are developing new ways to reuse and recycle the world’s products and materials. Our goal is to make circularity so easy and economical that we no longer generate waste or extract raw resources that don’t have regenerative properties.

_{Inside the lab, our technology scans pieces of plastic packaging and identifies their molecular composition in real-time.}

The problem with plastics

A biochemist by training, I’ve spent the majority of my career working at the intersection of sustainability and materials design—my last company developed a new material that was used in the first recyclable windmill blade. I knew that in order to tackle such a massive, multifaceted challenge as materials circularity, the team needed to begin by homing in on one specific area.

We decided to start with plastics. Since the 1950s, humanity’s use of plastic has grown exponentially: it’s an important material that allows the modern world to work. It’s affordable, lightweight, durable, versatile, and used in everything from cars and planes to the packaging that keeps our food from spoiling.

But the world’s ubiquitous plastic consumption has also become an environmental hazard when mismanaged: plastics account for 11 percent of the world’s global oil demand, and the world is producing twice as much plastic waste as we were two decades ago.

Our waste management systems were initially designed for glass and cardboard, and are still trying to catch up to the rapidly evolving use of complex plastic packaging. Modern packaging is often composed of many different plastic polymers and additives. When combined with other materials like paper and metal, our traditional waste facilities have a hard time identifying and routing them into proper recycling streams.

Black plastics, like the containers for takeout food, are particularly challenging because they’re invisible against the black conveyor belts inside recycling centers, where most cameras are incapable of identifying them. Flexible plastics such as granola pouches and potato chip bags have highly complex compositions that today are too complicated for current recycling systems.

In the U.S. alone, there are more than 9,000 unique recycling programs, with rules that vary widely. Certain communities accept things that others don’t, and tossing anything that seems recyclable into the blue bin creates a phenomenon known as “wishcycling.”

Today’s plastic recycling is mainly successful when the packaging is primarily made up of one molecular component, such as a polymer like polyethylene terephthalate (PET), commonly found in single-use water bottles, or like polyethylene (PE), found in milk jugs. The majority of the rest of the plastic the world produces ends up in landfills.

Analyzing waste at the molecular level

As our team began exploring different approaches to our circularity moonshot, we realized that most of the issues affecting today’s recycling industry stem from a lack of data. In order to recycle plastic materials with differing and complex chemical properties, we first must be able to easily identify what, exactly, they’re composed of.

_{Breaking down waste into its raw, chemical components could allow us to create new, everyday products without extracting resources from the earth.}

Today’s waste management industry has access to very poor, if any, data about the items it collects. Little information exists about the quality, quantity, and location of plastic waste. And the industry has no data about plastic waste at the molecular or chemical level; when our team tried to buy it for our research purposes, all we received were scanned printouts with very sparse information.

We realized that we needed to create this dataset ourselves. 

Using a combination of machine learning, AI, and Google’s world-class compute power, we’re building the industry’s first comprehensive database to catalog plastics in packaging. We started four years ago with our own household garbage, and even enlisted our families to cut out the samples from every package of food we bought. We used spectrometers and analytical chemistry to reveal the molecular composition of each package, and trained our algorithms how to properly identify the molecular makeup of each yogurt container or take out container.

Today, we have millions of data points about the everyday packaging we all use. We’re currently running a pilot project at a recycling plant in Oregon, where our technology scans thousands of pieces of plastic packaging and containers flying along massive conveyor belts every minute, and then instantly identifies their molecular composition, processing the data in real-time at high speed.

Our team’s unique sensor fusion system has the ability to determine the detailed, molecular makeup of household packaging, including black plastics, and we’re training our models to identify molecular components in other waste materials. We’re also developing new sorting processes to improve mechanical recycling. You can take a peek at what we’ve been up to in a recent episode of “Where the Internet Lives."

_{Watch the latest episode of "Where the Internet Lives" to learn more about the Moonshot for Circularity tools and technology.}

Our goal is to create an entirely new system in which waste management facilities can identify and sort all materials based on their molecular components. We believe that understanding waste materials at the molecular level will be the key to unlocking recycling and circularity at a much larger scale.

_{Identifying and sorting materials in the lab.}

Transforming trash into treasure

Beyond recycling, our moonshot extends to all aspects of the circular economy, including reuse and redesign.

If we can identify materials such as textiles at the molecular level and determine the quality is still adequate, we can route them back into the world for reuse, and minimize the need to even recycle or remanufacture the materials. And by knowing all the components that go into certain complex, impossible-to-recycle materials, we may be able to influence the way manufacturers design their products in the first place.

Longer term, the ability to identify the molecular components of waste also creates a massive opportunity to produce virgin-quality materials from these resources, advancing new approaches in “molecular” or “chemical” recycling. Successfully breaking down waste materials into their raw, chemical components in order to create new everyday products could help humanity wean itself off extracting petroleum and other resources from the earth entirely.

If you’re interested in collaborating or learning more, reach out to circularitymoonshot@google.com.