If you shop on Google Flights, you get a quick comparison for different itineraries: One flight’s carbon emissions may be average, while another’s are 14% higher. But if you go shopping for a new laptop, you likely won’t find quick, comprehensible information on different models’ sustainability bonafides, despite the of producing and discarding electronics. In part, that’s because understanding a device’s emissions is difficult and time-consuming, even for experts.��

����ԭ�� researchers developed an artificial intelligence system that automatically estimates the environmental impacts of making different electronic devices. The system uses AI agents — programs that perform tasks autonomously — to comb through publicly available data and conduct life cycle assessments, or LCAs. The system achieves an average error rate of 5%-19%, similar to the accuracy of LCAs conducted by experts.

The team June 12 in Nature Electronics.��

AI agents have recently grown increasingly capable of performing complex tasks. Today’s agents can search the web and pull information about electronic parts from product descriptions, images and documents.��

“Some of our previous research made me curious about how LCA experts perform environmental assessments — and whether that process could be automated,” said lead author , a UW doctoral student in the Allen School. “So to understand the bottlenecks firsthand, and then built a system that emulates these interactions with two AI agents. Each of them mimics different roles in the LCA process.”

One agent acts as a sort of analyst, defining what information needs to be gathered and how it will fit together. It also reviews results for accuracy. The second agent is more like an engineer. It scrapes publicly available data for information on an electronic device’s components. That might entail sifting through spreadsheets, or looking up images of the insides of devices and taking chip information from them — including from sources not typically used for LCAs, such as and posts on.��

The two agents work in a loop. The first sets the scope, the second gathers information. The first then looks that information over and might send the second agent searching again, and so on. The agents then reference to convert the complete list of parts to carbon estimates.

The team also developed a new method to bypass this detailed data collection and directly estimate carbon footprints. For common devices like laptops and smartphones with publicly available carbon footprint reports, they found that products with similar specs like screen size and processors clustered around similar carbon values, because only a handful of companies make specialized parts for all these devices. So an unknown device’s footprint can be represented as a weighted average of similar products.��

They also use this to estimate the carbon for materials not in LCA databases. For example, a new type of sustainable plastic could be estimated based on plastics with similar properties and chemistry.

“We tried this ‘nearest-neighbors’ approach and found that for materials, it’s actually better than the standard approach of a human picking the single closest entry,” said Zhang. “When estimating missing emissions factors in a test, the average error for our method was 23%. Human experts had an average error of 143%.” 

The authors note that while the aim of the system is to help reduce carbon emissions overall, running AI models requires energy, so they’ve taken several steps to mitigate its impact. They use small AI models that aren’t as energy-intensive as general-purpose models. They also start the process by running a search to see if the device’s estimated emissions have already been calculated. If so, it can stop there. If the system does need to call its AI models repeatedly, estimating a device’s carbon footprint is currently on par with the emissions generated by brewing a cup of tea.

The team plans to collaborate with companies in the future to help automate their workflows.��

“A lot of big companies have sustainability teams that perform these LCAs,” Iyer said. “Our hope is that automating this will actually free up their time, so they can spend their time reducing the carbon footprint of the products themselves, instead of hunting down elusive stats.” 

Co-authors include , a UW student in the Allen School;, , a UW postdoctoral researcher in the Allen School; , a UW doctoral student in the Allen School; , a UW professor in the Allen School; of Wesleyan University, who completed this research as a UW doctoral student in the Allen School; of the University of Notre Dame; of Northeastern University; and of Brown University, who completed this research as a UW assistant professor in the Allen School.��

This research was funded by Amazon Research Awards and the National Science Foundation. Zhang was supported by the .

For more information, contact Iyer at vsiyer@uw.edu and Zhang at zzhihan@cs.washington.edu.

A recent found that the world generated 137 billion pounds of electronic waste in 2022, an 82% increase from 2010. Yet less than a quarter of 2022’s e-waste was recycled. While many things impede a sustainable afterlife for electronics, one is that we don’t have systems at scale to recycle the found in nearly all electronic devices.

PCBs — which house and interconnect chips, transistors and other components — typically consist of layers of thin glass fiber sheets coated in hard plastic and laminated together with copper. That plastic can’t easily be separated from the glass, so PCBs often pile up in landfills, where their chemicals can seep into the environment. Or they’re burned to extract their electronics’ valuable metals like gold and copper. This burning, , is wasteful and can be toxic — especially for those doing the work without proper protections.

A team led by researchers at the ����ԭ�� developed a new PCB that performs on par with traditional materials and can be recycled repeatedly with negligible material loss. Researchers used a solvent that transforms a type of — a cutting-edge class of sustainable polymers — to a jelly-like substance without damaging it, allowing the solid components to be plucked out for reuse or recycling.

The vitrimer jelly can then be repeatedly used to make new, high-quality PCBs, unlike conventional plastics that degrade significantly with each recycling. With these “vPCBs” (vitrimer printed circuit boards), researchers recovered 98% of the vitrimer and 100% of the glass fiber, as well as 91% of the solvent used for recycling.

The researchers published April 26 in Nature Sustainability.

“PCBs make up a pretty large fraction of the mass and volume of electronic waste,” said co-senior author , a UW assistant professor in the Paul G. Allen School of Computer Science & Engineering. “They’re constructed to be fireproof and chemical-proof, which is great in terms of making them very robust. But that also makes them basically impossible to recycle. Here, we created a new material formulation that has the electrical properties comparable to conventional PCBs as well as a process to recycle them repeatedly.”

Vitrimers are a class of polymers first developed in 2015. When exposed to certain conditions, such as heat above a specific temperature, their molecules can rearrange and form new bonds. This makes them both “healable” (a bent PCB could be straightened, for instance) and highly recyclable.

“On a molecular level, polymers are kind of like spaghetti noodles, which wrap and get compacted,” said co-senior author , a UW assistant professor in the mechanical engineering department. “But vitrimers are distinct because the molecules that make up each noodle can unlink and relink. It’s almost like each piece of spaghetti is made of small Legos.”

The team’s process to create the vPCB deviated only slightly from those used for PCBs. Conventionally, semi-cured PCB layers are held in cool, dry conditions where they have a limited shelf life before they’re laminated in a heat press. Because vitrimers can form new bonds, researchers laminated fully cured vPCB layers. The researchers found that to recycle the vPCBs they could immerse the material in an organic solvent that has a relatively low boiling point. This swelled the vPCB’s plastic without damaging the glass sheets and electronic components, letting the researchers extract these for reuse.

This process allows for several paths to more sustainable, circular PCB lifecycles. Damaged circuit boards, such those with cracks or warping, can in some cases be repaired. If they aren’t repaired, they can be separated from their electronic components. Those components can then be recycled or reused, while the vitrimer and glass fibers can get recycled into new vPCBs.

The team tested its vPCB for strength and electrical properties, and found that it performed comparable to the most common PCB material (). Vashisth and co-author , a principal researcher at Microsoft Research and an affiliate assistant professor in the Allen School, are now using artificial intelligence to explore new vitrimer formulations for different uses.

Producing vPCBs wouldn’t entail major changes to manufacturing processes.

“The nice thing is that a lot of industries — such as aerospace, automotive and even electronics — already have processing set up for the sorts of two-part epoxies that we use here,” said lead author , a UW doctoral student in the Allen School.

The team analyzed the environmental impact and found recycled vPCBs could entail a 48% reduction in global warming potential and an 81% reduction in carcinogenic emissions compared to traditional PCBs. While this work presents a technology solution, the team notes that a significant hurdle to recycling vPCBs at scale would be creating systems and incentives to gather e-waste so it can be recycled.

“For real implementation of these systems, there needs to be cost parity and strong governmental regulations in place,” said Nguyen. “Moving forward, we need to design and optimize materials with sustainability metrics as a first principle.”

Additional co-authors include , a UW postdoctoral scholar in the mechanical engineering department; , a UW doctoral student in the mechanical engineering department; , a senior applied scientist at Microsoft Research; , a senior researcher at Microsoft Research and an affiliate researcher in the Allen School; and , a UW professor in the Allen School and the electrical and computer engineering department. This research is funded by the Microsoft Climate Research Initiative, an Amazon Research Award and the Google Research Scholar Program. Zhang was supported by the UW Clean Energy Institute Graduate Fellowship.

For more information, contact vpcb@cs.washington.edu.