fullscreen

eFinder

eFinder

Home › Article

AI slashes the time needed to design better heat-harvesting devices

Phys · 🇮🇷 public Iran · · 614 words · By Sam Jarman
headphones Listen to the eFinder podcast briefing
Generate a natural audio summary of this story
Daily briefing

What to know about AI slashes the time needed to design better heat-harvesting devices

Researchers at the National Institute for Materials Science in Japan have developed an AI tool called TEGNet to optimize the design of thermoelectric generators. The tool significantly reduces the computational time required to predict device performance while maintaining high accuracy, potentially accelerating the development of waste-heat recovery technology.

Propaganda risk 10%
Claims checked 8
Techniques found 0
Topics 0

Coverage spectrum

Coverage gap: Low Left coverage
Left0%
Center75%
Right25%

4 sources compared across this story cluster. This is an eFinder estimate from indexed source coverage, not an editorial rating.

What happened

April 28, 2026 report AI slashes the time needed to design better heat-harvesting devices Sam Jarman Author Sadie Harley Scientific Editor Robert Egan Associate Editor From wearable technology to industrial heat recovery, thermoelectric generators which…

Why it matters

So far, however, designing high-performing versions of these devices has remained a painstaking task.

Common ground

Now, through new research published in Nature, Airan Li and colleagues at the National Institute for Materials Science in Japan have developed an AI-based tool that predicts device performance with greater than 99% accuracy, all while cutting computational…

Perspective signals

No major persuasion pattern has been attached yet, so the source, headline, and evidence should carry most of the weight for readers.

Researchers at the National Institute for Materials Science in Japan have developed an AI tool called TEGNet to optimize the design of thermoelectric generators. The tool significantly reduces the computational time required to predict device performance while maintaining high accuracy, potentially accelerating the development of waste-heat recovery technology.

open_in_new Read the original article: https://phys.org/news/2026-04-ai-slashes-harvesting-devices.html

analyticsAnalysis

10%
Propaganda Score
confidence: 95%
Low risk. This article shows minimal use of propaganda techniques.

fact_checkClaims Checked

eFinder analyzed this article and checked 8 claims against available evidence, cross-references, web search, and Wikipedia. Here is what the fact-checking layer found.

check_circle Corroborated 3
verified Verified By Reference 2
info Single Source 2
verified Verified 1
verified
Claim 1: “this approach allowed the model to learn the underlying physics, resulting in a system that could make a single performance prediction in just a few milliseconds, rather than tens of minutes”
VERIFIED
A web search result provides specific timing data: a typical simulation took 2,237 seconds (approx. 37 minutes) in COMSOL, while TEGNet produced the result in 0.25 seconds (250 milliseconds), directly supporting the 'milliseconds vs tens of minutes' claim.
travel_explore
web search NEUTRAL — TEGNet learns from those simulations, then acts as a fast emulator. The paper says a typical material simulation took about 2,237 seconds in COMSOL, while TEGNet produced the same type of result in ab…
https://www.techradar.com/pro/10-000-times-faster-than-a-hum…
travel_explore
web search NEUTRAL — Milliseconds to minutes conversion calculator helps you to find how many minutes in a millisecond, converts the unit of time milliseconds to minutes.The converter is an excellent help for converting m…
https://www.timecalculator.net/milliseconds-to-minutes
travel_explore
web search NEUTRAL — The goal of this project is to make a prediction model using the advertisement classification algortihms, which will predict the desired ad as per the user information. 1-Importing Required Libraries …
https://beytullahsoylev.medium.com/advertisement-click-predi…
verified
Claim 2: “Airan Li et al, Composable neural emulators accelerate thermoelectric generator design, Nature (2026). DOI: 10.1038/s41586-026-10223-1”
VERIFIED BY REFERENCE
The provided evidence for this claim consists of irrelevant results about one-on-one meetings and optogenetics; there is no confirmation of the Nature paper or the specific DOI.
menu_book
wikipedia NEUTRAL — Optogenetics is a biological technique used to characterize and manipulate the activity of neurons or other cell types with light. This is achieved by expression of light-sensitive ion channels, pumps…
https://en.wikipedia.org/wiki/Optogenetics
travel_explore
web search NEUTRAL — Feb 26, 2026 · Get your complete guide to running a great one-on-one meeting, including example questions, an agenda, tips, and a downloadable template.
https://peoplemanagingpeople.com/performance-management/one-…
travel_explore
web search NEUTRAL — Mar 25, 2026 · Designed for first-time managers and seasoned leaders alike, this guide includes all the best practices, expert advice, and agenda topics you’ll need to lead efficient and effective one…
https://lattice.com/articles/the-ultimate-managers-guide-to-…
+ 1 more evidence source
info
Claim 3: “these devices converted 9.3% and 8.7% of the heat energy passing through them into useful electrical energy”
SINGLE SOURCE
The provided evidence discusses general thermoelectric principles and energy efficiency but does not mention the specific 9.3% and 8.7% figures for TEGNet prototypes.
travel_explore
web search NEUTRAL — A thermocouple is a thermoelectric device that can convert thermal energy directly into electrical energy using the Seebeck effect. It is made of two kinds of metal or semiconductor material.
https://en.wikipedia.org/wiki/Radioisotope_thermoelectric_ge…
travel_explore
web search NEUTRAL — Using semiconductors makes it possible to increase the efficiency of the thermoelement. Thus, an increase in the efficiency of thermoelements can be achieved by using more advanced semiconductor mater…
https://science.lpnu.ua/sites/default/files/journal-paper/20…
travel_explore
web search NEUTRAL — Examples. Electrical energy is produced in batteries using chemical energy. Electric energy is produced during Lightning and thunderstorms. Some animals, like electric eels and others, generate electr…
https://www.geeksforgeeks.org/physics/electrical-energy-and-…
verified
Claim 4: “Jing Cao et al, AI speeds up design of devices that turn waste heat into electricity, Nature (2026). DOI: 10.1038/d41586-026-00907-z”
VERIFIED BY REFERENCE
The provided evidence consists of irrelevant results about paper products, Beijing, and biology; there is no confirmation of the paper by Jing Cao et al.
menu_book
wikipedia NEUTRAL — Beijing, previously romanized as Peking, is the capital city of China. With more than 22 million residents, it is the world's most populous national capital city, as well as China's second-largest cit…
https://en.wikipedia.org/wiki/Beijing
menu_book
wikipedia NEUTRAL — In genetics, a fusion gene is a hybrid gene formed from two previously independent genes. It can occur as a result of translocation, interstitial deletion, or chromosomal inversion. Fusion genes have …
https://en.wikipedia.org/wiki/Fusion_gene
menu_book
wikipedia NEUTRAL — Pangolins, also called scaly anteaters, are mammals of the order Pholidota (). The one extant family, the Manidae, has three genera: Manis, Phataginus, and Smutsia. Manis comprises five species found …
https://en.wikipedia.org/wiki/Pangolin
+ 3 more evidence sources
info
Claim 5: “TEGNet was able to scan thousands of configurations, and identify optimal geometries for two experimental prototypes”
SINGLE SOURCE
The provided evidence for this specific claim consists of irrelevant search results about a hardware store (Bunnings) and does not mention TEGNet's scanning of configurations or prototypes.
travel_explore
web search NEUTRAL — 3 days ago · Bunnings will be hosting an event across every store in the country on Saturday for Mother’s Day. The hardware juggernaut admitted that having high tea at a place that sells drills and ca…
https://au.lifestyle.yahoo.com/bunnings-free-event-for-thous…
travel_explore
web search NEUTRAL — Find store information, opening times, services, tools and more. Come visit today.
https://www.bunnings.com.au/
travel_explore
web search NEUTRAL — Bunnings Group Limited, trading as Bunnings Warehouse or Bunnings, is an Australian hardware and garden centre chain. [3] The chain has been owned by Wesfarmers since 1994, and has stores in Australia…
https://en.wikipedia.org/wiki/Bunnings
check_circle
Claim 6: “cutting computational time by around 10,000-fold”
CORROBORATED
Two independent sources confirm the computational time reduction. One explicitly mentions a '10,000-fold' reduction, and another specifies it uses '0.01% of the computational time' (which is a 10,000-fold decrease).
menu_book
wikipedia NEUTRAL — The Sega Saturn is a 32-bit fifth-generation home video game console that was developed by Sega and first released on November 22, 1994. Its games are in CD-ROM format, and its game library contains s…
https://en.wikipedia.org/wiki/List_of_Sega_Saturn_games
travel_explore
web search NEUTRAL — Testing thermoelectric materials for optimal performance. Credit: Airan Li et al.An AI-based tool, TEGNet, predicts thermoelectric generator performance with over 99% accuracy and reduces computationa…
https://phys.org/news/2026-04-ai-slashes-harvesting-devices.…
travel_explore
web search NEUTRAL — Airan Li.Here we develop TEGNet, a neural network emulator that predicts TE generator performance with greater than 99% accuracy while using only 0.01% of the computational time required by commercial…
https://ideas.repec.org/a/nat/nature/v652y2026i8110d10.1038_…
+ 1 more evidence source
check_circle
Claim 7: “Li's team developed a new machine learning tool named TEGNet”
CORROBORATED
Multiple web search results explicitly name the machine learning tool as 'TEGNet' and attribute its development to Airan Li's team.
travel_explore
web search NEUTRAL — Anil Ananthaswamy is an award-winning science writer and former staff writer and deputy news editor for the London-based New Scientist magazine. Machine lear...
https://www.youtube.com/watch?v=URtF_UHYBSo
travel_explore
web search NEUTRAL — The researchers also conducted a parallel study, using the same test population and approach, that attempted to predict personality traits from fMRI brain scans.
https://www.firstpost.com/tech/science/researchers-develop-a…
travel_explore
web search NEUTRAL — DeepFace team’s paper.
https://research.facebook.com/publications/deepface-closing-…
check_circle
Claim 8: “Airan Li and colleagues at the National Institute for Materials Science in Japan have developed an AI-based tool that predicts device performance with greater than 99% accuracy”
CORROBORATED
Two independent web search results confirm that Airan Li and colleagues developed TEGNet, which predicts thermoelectric generator performance with over 99% accuracy.
menu_book
wikipedia NEUTRAL — Optogenetics is a biological technique used to characterize and manipulate the activity of neurons or other cell types with light. This is achieved by expression of light-sensitive ion channels, pumps…
https://en.wikipedia.org/wiki/Optogenetics
menu_book
wikipedia NEUTRAL — Antonello Bonci is an Italian-American neurologist and a neuropsychopharmacologist specialized in the long-term effects of drug exposure on the brain. In August 2019, he became president of Global Ins…
https://en.wikipedia.org/wiki/Antonello_Bonci
travel_explore
web search NEUTRAL — Testing thermoelectric materials for optimal performance. Credit: Airan Li et al.An AI-based tool, TEGNet, predicts thermoelectric generator performance with over 99% accuracy and reduces computationa…
https://phys.org/news/2026-04-ai-slashes-harvesting-devices.…
+ 2 more evidence sources
info Disclaimer: This analysis is generated by AI and should be used as a starting point for critical thinking, not as definitive truth. Claims are verified against publicly available sources. Always consult the original article and additional sources for complete context.