Sustainability assessment using multimodal artificial intelligence agents
- 1Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, WA, USA
- 2Computer Science and Engineering, University of Notre Dame, Notre Dame, IN, USA
- 3Electrical and Computer Engineering, Northeastern University, Boston, MA, USA
Abstract
Reducing the growing environmental impact of the computing industry requires assessing the emissions of electronics at scale. However, a traditional life-cycle assessment (LCA) of an electronic device, which maps materials and processes to environmental impacts, often requires proprietary or unavailable data. Here we report a multimodal multi-agent artificial intelligence system that emulates the collaborative process between LCA professionals and stakeholders (such as product managers and engineers) to estimate the carbon footprint of electronic devices. The agents iteratively construct a complete life-cycle inventory by leveraging a structured data abstraction and software tools that mine information from the public Internet, including repair communities and government regulatory databases. This reduces data gaps and data collection from weeks or months of expert time to under 1 min. The system can calculate the carbon footprint within 19% of expert LCAs with zero proprietary data (typical of the variation between human LCAs). We also show that by encoding domain-specific knowledge, environmental impact estimation can be reframed as a data-driven prediction task, in which both unknown products and emission factors are represented as weighted combinations of similar ones with known emissions. A multimodal multi-agent artificial intelligence system that emulates the collaborative process between life-cycle assessment professionals and stakeholders (such as product managers and engineers) can calculate the carbon footprint of electronic devices, performing data collection tasks that can take weeks or months of expert time to under a minute.
Related
- Artificial intelligence-enabled monitoring of pesticide contamination in agricultural water systems: a comprehensive review of intelligent sensing technologies for water quality surveillance and environmental risk assessment
- Global food security under a sustainability perspective: Assessment, challenges, and strategies
- Evaluation of sustainability of lubricants by life cycle assessment and carbon footprint measurement
- Carbon footprint of biomass-derived activated carbon: life cycle assessment, environmental sustainability, and carbon-negative potential.
- Clarifying environmental responsibility for consistent absolute environmental sustainability assessment: A global sector-level analysis
BibTeX
@article{zhang2026sustainability,
title = {{Sustainability assessment using multimodal artificial intelligence agents}},
author = {Zhihan Zhang and Alexander Le Metzger and Yuxuan Mei and Felix Hähnlein and Zachary Englhardt and Tingyu Cheng and Gregory D. Abowd and Shwetak Patel and Adriana Schulz and Vikram Iyer},
journal = {Nature Electronics},
year = {2026},
month = jun,
doi = {10.1038/s41928-026-01653-w},
url = {https://doi.org/10.1038/s41928-026-01653-w},
note = {CC BY},
}
RIS
TY - JOUR
TI - Sustainability assessment using multimodal artificial intelligence agents
AU - Zhihan Zhang
AU - Alexander Le Metzger
AU - Yuxuan Mei
AU - Felix Hähnlein
AU - Zachary Englhardt
AU - Tingyu Cheng
AU - Gregory D. Abowd
AU - Shwetak Patel
AU - Adriana Schulz
AU - Vikram Iyer
PY - 2026
DA - 2026/06/12
JO - Nature Electronics
DO - 10.1038/s41928-026-01653-w
UR - https://doi.org/10.1038/s41928-026-01653-w
AB - Reducing the growing environmental impact of the computing industry requires assessing the emissions of electronics at scale. However, a traditional life-cycle assessment (LCA) of an electronic device, which maps materials and processes to environmental impacts, often requires proprietary or unavailable data. Here we report a multimodal multi-agent artificial intelligence system that emulates the collaborative process between LCA professionals and stakeholders (such as product managers and engineers) to estimate the carbon footprint of electronic devices. The agents iteratively construct a complete life-cycle inventory by leveraging a structured data abstraction and software tools that mine information from the public Internet, including repair communities and government regulatory databases. This reduces data gaps and data collection from weeks or months of expert time to under 1 min. The system can calculate the carbon footprint within 19% of expert LCAs with zero proprietary data (typical of the variation between human LCAs). We also show that by encoding domain-specific knowledge, environmental impact estimation can be reframed as a data-driven prediction task, in which both unknown products and emission factors are represented as weighted combinations of similar ones with known emissions. A multimodal multi-agent artificial intelligence system that emulates the collaborative process between life-cycle assessment professionals and stakeholders (such as product managers and engineers) can calculate the carbon footprint of electronic devices, performing data collection tasks that can take weeks or months of expert time to under a minute.
SN - 2520-1131
ER -