Pith. sign in

REVIEW 4 major objections 5 minor 64 references

Sustainability assessment using multimodal AI agents

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A multi-agent AI system estimates the carbon footprint of electronic devices within 19% of expert life cycle assessments using only public data.

desk verdict Solid end-to-end system with real results, but the '19% of expert LCAs' headline overstates a benchmark measured against company-reported PCFs, not expert LCAs. read the letter →

arxiv 2507.17012 v2 pith:XGRPLRQL submitted 2025-07-22 cs.AI cs.CE

Zhihan Zhang , Alexander Metzger , Yuxuan Mei , Felix Hähnlein , Zachary Englhardt , Tingyu Cheng , Gregory D. Abowd , Shwetak Patel
show 2 more authors
Adriana Schulz Vikram Iyer
This is my paper · ORCID
classification cs.AIcs.CE
keywords lifecycleassessmentcarbonfootprintestimationmulti-agentAImultimodalretrievalconsumerelectronicsemissionfactorsweightedk-nearestneighborssustainablecomputing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that a full life cycle assessment of a consumer electronic device can be automated end to end: give the system a product name or photo, and paired AI agents build a component-level inventory from public photos and documents, then translate it into a carbon footprint using standard emission factors. The aim is to remove the proprietary-data bottleneck that makes conventional LCA slow, expensive, and hard to verify. The system reports estimates within 19% of expert LCAs on iPhones and across seven unseen product categories, with a mean absolute percentage error of 18.02% on iPhones, and it reframes impact estimation as a data-driven prediction task in which products and emission factors are approximated by weighted combinations of similar known entries. A reader should care because this is the accuracy regime that makes rapid early-design guidance and transparent public disclosure practical rather than expert-only analysis.

What carries the argument

The load-bearing mechanism is the multi-agent self-play loop organized by a structured data abstraction. The LCA agent builds a data abstraction listing component classes (PCB, ICs, sensors, passives, mechanical parts, battery, display) that constrains the inventory to methodologically valid system boundaries; it then critiques completeness and issues targeted queries. The stakeholders agent answers by orchestrating tools: web search for text and images, an FCC database scraper for internal photos, a vision toolchain of FFT Gaussian high-pass filtering and a YOLO v11 detector for PCB selection and component counting, OCR and SAM segmentation for dimensions using visible scale references, and branched retrieval for part numbers and process technology. Standard LCIA multiplies the completed inventory by emission factors from ecoinvent and related LCA literature. A second mechanism, the weighted k-nearest-neighbors Gaussian estimator, carries the direct-prediction results: it finds neighbors by Euclidean distance in a domain-feature space, weights them by attribute completeness, returns the mean as a point estimate, and reports the neighbor spread as uncertainty.

What would settle it

Take a set of devices whose manufacturers publish independently audited, primary-data life cycle assessments with full supply chain data rather than public PCF summaries, run the agentic pipeline on each, and compute the mean absolute percentage error between predicted and audited cradle-to-gate CO2e; the 19% claim fails if that error exceeds 19% or if error patterns track the opacity of public teardown data.

Watch

Extended reading notes

Core claim

The paper's central claim is that an end-to-end life cycle assessment for consumer electronics no longer requires proprietary bills of materials or months of expert labor. Given only a product name or photo, two large-language-model agents play a self-play game: an LCA agent defines a data abstraction that fixes the system boundary and iteratively critiques the growing inventory, while a stakeholders agent retrieves product specifications, FCC internal photos, and community teardowns, then uses a vision pipeline (FFT filtering, a YOLO component detector, OCR, and segmentation with ruler calibration) to estimate component categories and dimensions. The resulting component-level inventory is multiplied by standard emission factors to produce a cradle-to-gate carbon footprint. The authors report this end-to-end system matches Apple's reported iPhone carbon footprints with an MAPE of 18.02%, stays within 5-19% of reported values across seven product categories not seen during development, and matches expert LCAs within 19% overall, a margin the paper notes is typical of human LCA variation. The same pipeline reframes environmental impact estimation as a learnable prediction task: a weighted k-nearest-neighbor estimator over domain features (technology node, memory capacity, display type, material properties) maps unknown products and unknown emission factors to weighted combinations of similar known entries, with reported MAPEs of 12.28% within a company and 16.54% across companies after calibration, and better emission-factor matches than nine human experts.

Load-bearing premise

The whole accuracy claim is anchored to company-reported product carbon footprints from Apple, Asus, Dell, HP, and Lenovo being a reliable stand-in for expert LCA, even though the paper itself says these reports are not ground truth because their system boundaries and assumptions are opaque.

Editorial extensions

If this is right

  • Product designers can get a carbon estimate for a new device from a name or photo in under a minute, making hotspot identification practical early in design rather than after months of expert LCA work.
  • Companies and regulators can benchmark reported product carbon footprints against inventories derived from public regulatory filings and repair teardowns, increasing transparency and reproducibility of sustainability claims.
  • With roughly 80-120 labeled examples, the weighted k-NN estimator reaches practical accuracy, so the approach can be deployed in data-sparse LCA domains beyond electronics.
  • Inference-time scaling shows that longer thinking time, more retrieved documents, and more reasoning steps improve accuracy with diminishing returns beyond about 40 seconds, giving users a concrete accuracy-versus-compute trade-off.
  • Automated retrieval and neighbor-based estimation can flag environmental hotspots early enough to guide design changes, even when absolute accuracy is insufficient for formal carbon accounting.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If company-reported product carbon footprints are themselves methodologically inconsistent, the claimed 19% accuracy is really a match to those reported numbers rather than to an absolute environmental truth; true error could be larger or smaller.
  • The same agentic retrieval design should transfer to other product classes with public teardown or regulatory-photo cultures, such as automobiles, appliances, or medical devices, extending the method beyond electronics.
  • The neighbor-based estimator's uncertainty intervals could serve as defensible priors for regulatory reporting or for deciding where to invest in primary data collection, concentrating audit effort on products with high neighbor variance.
  • Linking the agentic inventory generator to generative design tools could give designers immediate carbon feedback per design alternative, a natural extension of the paper's rapid-estimation goal.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The manuscript presents a multimodal multi-agent AI system that automatically constructs life-cycle inventories (LCI) for electronic products from public sources (FCC filings, iFixit teardowns, product pages) and estimates cradle-to-gate carbon footprints. A second contribution is a weighted k-nearest-neighbor estimator that predicts product carbon footprints from domain-specific textual features, and a third is a data-driven method for estimating emission factors for LCA database entries. The central claims are that the end-to-end system estimates product carbon footprints within 19% of 'expert LCAs' with zero proprietary data, that the k-NN estimator reaches a MAPE of 12.28% within a company and 16.54% cross-company after calibration, and that the emission-factor estimator outperforms human experts. Evaluations are conducted against company-reported product carbon footprints (Apple, Asus, Dell, HP, Lenovo) and against ecoinvent emission factors, with human-expert benchmarking for the emission-factor task.

Significance. If the 19% accuracy claim were established against genuinely independent expert LCAs, this would be a significant contribution: automated, fast, zero-proprietary-data carbon footprint estimation could support early-stage design and disclosure workflows. The paper has clear strengths: an end-to-end pipeline from a product name, use of previously untapped public data sources, a well-structured multi-agent architecture with tool orchestration, a user study, and public code and data availability statements. The k-NN estimator with interpretable nearest neighbors is a sensible alternative to black-box models for LCA, and the human-benchmark comparison for emission-factor matching is a valuable methodological addition. However, the headline accuracy claim is measured against company-reported PCFs, which the paper itself concedes are not ground truth, so the stated equivalence to 'expert LCAs' is not yet established. The evaluation sample is also restricted to products with public teardown imagery and published PCF reports, which limits the generality of the end-to-end claim.

major comments (4)
  1. [Abstract, §1, §2 (Fig. 2d/f), §6, §7.3] The headline claim that the system 'can calculate carbon footprint within 19% of expert LCAs' is not supported by the evaluation as designed. The MAPE values are computed against company-reported product carbon footprints (Apple, Asus, Dell, HP, Lenovo), and §7.3 explicitly states that these reports 'are not ground truth, due to a lack of transparency in underlying system boundaries and assumptions.' Section 6 further acknowledges that there is 'no unambiguously known true value' for an LCA. Since the Introduction itself notes that human expert LCAs can vary by 20% or more (and up to threefold), agreement with one company's self-reported figure does not establish agreement with an independent expert LCA. This is a load-bearing issue for the central claim. I recommend that the wording throughout—including the Abstract, Introduction, and Conclusions—be revised to 'within 19% of company-reported product carbon footprints,' and that the authors either add a subset validation against independent LCAs with fixed system boundaries or explicitly reframe the claim as benchmarking against corporate disclosures.
  2. [§2, §7.2, §7.3, Fig. 2f] The end-to-end evaluation is performed only on products for which public teardown/FCC photos and published PCF reports are available. The stated capability 'requires only a product name as input' (§1) is therefore not demonstrated for products that lack such disclosures, and the paper does not report the success rate or selection bias of the retrieval pipeline across a broader product population. The figure for cross-category generalization (Fig. 2f) shows only two products, while the text cites seven held-out categories. I request explicit reporting of the number of products per category, the total evaluation set size, and confidence intervals for the headline MAPEs, along with an analysis of how the pipeline behaves when FCC images are confidential and iFixit teardowns are sparse. Without this, the 5–19% range cannot be assessed for statistical reliability.
  3. [§3, Fig. 3g-j] The cross-company generalization result for the k-NN estimator depends on a 'distributional calibration' step that appears to require knowledge of the target company's aggregate PCF distribution. If this calibration uses the target company's reported PCFs at inference time, then the claim of generalizing to unseen manufacturers without proprietary data is weakened, and the method is not a strict zero-shot estimator. The paper should state explicitly whether the calibration factor is computed from the target company's PCFs, report the uncalibrated cross-company MAPE, and justify the calibration procedure as a legitimate methodological choice (e.g., as prior knowledge about company-level system boundary conventions) rather than as a form of target leakage.
  4. [§4, §7.4.2, §7.5] For the emission-factor estimation task, the 25 material features are engineered using domain knowledge that may have been derived from the same ecoinvent classes used for evaluation. The manuscript does not specify whether the feature values (e.g., melting point, manufacturing temperature) are taken from the ecoinvent entries themselves or from independent sources, and whether cross-validation is performed across ecoinvent entries rather than tokens. This matters because if the features are correlated with the emission-factor target through the database structure, the comparison to human experts could be favorably biased. I ask the authors to clarify the provenance of the feature values and to report a cross-validated evaluation that removes any direct dependence on the target emission factors.
minor comments (5)
  1. [§1] 'mean average percentage error (MAPE)' should be 'mean absolute percentage error' (or the definition should be aligned with standard usage). The abbreviation MAPE is used consistently but the expansion appears only in the Introduction.
  2. [§7.4.1] The bullets 'Integrated Circuits' and 'Random-Access Memory' are not parallel with the other bullets; 'Random-Access Memory' appears in title case while others are sentence case. Additionally, the actual list item text for 'Random-Access Memory' is missing its final period.
  3. [Fig. 2f] The caption states that the GPU and motherboard examples 'demonstrat[e] the generalizability' across seven held-out categories, but showing only two products in the figure is insufficient to support that statement; please either include plots for all seven categories in an Extended Data figure or cite Supplementary Table 1 more prominently with per-category sample sizes.
  4. [§5] The scaling analysis treats thinking time, reasoning steps, and documents read as three independent axes, but these quantities are intrinsically correlated (longer thinking time naturally leads to more steps and more documents). The causal attribution of performance gains to any single axis should be softened, or the axes should be decoupled experimentally.
  5. [§2] The phrase 'achieved a 100% success rate corresponding to the product name' is vague; please specify what is being measured (e.g., correct retrieval of internal photos for the queried product) and report the number of products in that test set.

Circularity Check

1 steps flagged · score 3.0 of 10

One secondary cross-company prediction is partially forced by target-company aggregate calibration; the central agentic-LCA claim is independently benchmarked.

  1. fitted input called prediction [Section 3 ('Estimating EI from Domain-Specific Textual Features'), cross-company evaluation; Fig. 3g-j and surrounding text.]
    "When applied to Dell products using a model trained exclusively on Asus data, the estimator achieves a PCF MAPE of 16.54% after distributional calibration (Fig. 3g-j), comparable to within-company performance. ... Calibration based on overall distributional differences, which reflect variations in LCA methodologies across companies, improves cross-company estimation."

    The calibration is based on 'overall distributional differences' between companies, which are aggregate statistics of the same company-reported PCF values that Section 7.3 designates as the reference baseline. Fitting a company-level shift to those reported values and then computing MAPE against them forces the systematic component of the error toward zero by construction, so the 16.54% mostly reflects scatter around the calibrated mean rather than the model's ability to predict an unseen manufacturer from Asus data alone. The step is transparently labeled 'after distributional calibration,' but the claim that the estimator 'generalizes well beyond its training distribution' is partially circular because the calibration consumes the target-company labels used for evaluation.

full rationale

The paper's central contribution, the multi-agent LCI pipeline, is evaluated against externally reported product carbon footprints and ecoinvent emission factors, so the headline accuracy claim is not derived from the system's own outputs. The 'correct-by-construction' data abstraction is justified by a self-citation (ref [39]), but the 18.02% iPhone MAPE and the 5-19% range across seven held-out categories provide independent empirical support, so this self-citation is not load-bearing for the main result. The one genuinely circular step is the cross-company k-NN evaluation: the 16.54% Dell MAPE is obtained 'after distributional calibration' based on 'overall distributional differences' between companies, meaning the target-company aggregate PCFs used as the benchmark are also used to fit the calibration. This partially forces the reported generalization result. The broader concern that company-reported PCFs are not ground truth, explicitly acknowledged in Sections 6 and 7.3, is a benchmark-validity and correctness issue rather than a circularity, and does not by itself raise the circularity score.

Assumptions & free parameters 6 free parameters · 7 assumptions · 0 invented entities

The central claims depend on several domain assumptions and hand-chosen parameters; the most consequential are the validity of company-reported PCFs as benchmarks, the sufficiency of public teardown data, and the feature sets used by the k-NN models.

free parameters (6)
  • number of nearest neighbors k = not fully specified; k=5 in example, varied for electricity grids
    k is chosen by evaluation and affects MAPE; no single justified value in main text.
  • Gaussian HPF cutoff for PCB filtering = 32 cycles per pixel
    Hand-chosen heuristic to identify PCB images; location Fig 2c.
  • YOLO detection thresholds = confidence 0.3, IoU 0.5
    Set to reduce false negatives; location Methods Section 7.2.
  • distributional calibration factor = not reported numerically
    Used to shift Asus-trained estimates to Dell's aggregate PCF distribution (Section 3, Fig 3g-j); this is a per-target-company parameter and weakens out-of-distribution claims.
  • data availability-aware weights = not specified
    Weighting scheme deferred to Supplementary 4; hand-designed to handle missing attributes in k-NN.
  • feature sets (40 electronics, 25 material) = 40, 25
    Domain-specific attributes chosen by expert hypothesis, not learned; these are modeling choices that the central k-NN and emission factor results depend on.
assumptions (7)
  • domain assumption Company-reported PCFs are sufficiently accurate to serve as benchmark labels for expert LCA.
    All validation uses reported values from Apple/Asus/Dell/HP/Lenovo; Section 7.3 admits they are not ground truth.
  • domain assumption Public FCC teardown images and repair-community photos contain enough visual detail to reconstruct a representative LCI.
    The visual pipeline depends on FCC internal photos and iFixit teardowns (Section 2, Methods 7.2); failures occur for confidential recent devices.
  • domain assumption Electronics PCF is dominated by a small set of attributes (IC generation, memory, storage, battery, display, dimensions, energy) and these attributes transfer across brands.
    Used to justify the 40-feature k-NN model (Section 3, Section 7.4.1); if attributes do not capture intra-class variance, the estimator degrades.
  • domain assumption Ecoinvent emission factors are correct and representative for LCIA.
    The agent multiplies LCI entries by closest ecoinvent factors; errors in ecoinvent propagate to all estimates.
  • domain assumption The data abstraction system boundary matches conventional cradle-to-gate LCA.
    The DA defines correct-by-construction inventory scope (Section 2); this is a methodological choice inherited from prior work [39].
  • ad hoc to paper For cross-company estimation, the target company's aggregate PCF distribution is available or knowable at inference time.
    The calibration step in Section 3 uses aggregate distributional differences; without this, cross-company MAPE would be higher.
  • domain assumption Nearest-neighbor Euclidean distance in handcrafted feature space is a meaningful similarity metric for carbon footprint.
    The k-NN estimator relies on this for both products and materials (Section 3, Section 7.4).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sustainability assessment using multimodal AI agents." pith.science (2026). https://pith.science/paper/XGRPLRQL

@misc{pith2026250717012,
  author       = {Pith},
  title        = {Pith review of: Sustainability assessment using multimodal AI agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XGRPLRQL}},
  note         = {Machine review of arXiv:2507.17012}
}
read the original abstract

Reducing the rapidly 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 reimagine conventional sustainability assessment by introducing a multimodal multi-agent AI system that emulates the collaborative process between LCA professionals and stakeholders (such as product managers and engineers) to automatically 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 one minute. The system can calculate 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.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 46 canonical work pages

  1. [1]

    Patterns2(9), 100340 (2021) https://doi.org/10.1016/j

    Freitag, C., Berners-Lee, M., Widdicks, K., Knowles, B., Blair, G.S., Friday, A.: The real climate and transformative impact of ICT: A critique of estimates, trends, and regulations. Patterns2(9), 100340 (2021) https://doi.org/10.1016/j. patter.2021.100340 . Accessed 2023-11-13

    arXiv 2021

  2. [2]

    Challenges6(1), 117–157 (2015) https://doi.org/10.3390/ challe6010117

    Andrae, A.S.G., Edler, T.: On Global Electricity Usage of Communication Tech- nology: Trends to 2030. Challenges6(1), 117–157 (2015) https://doi.org/10.3390/ challe6010117 . Number: 1. Accessed 2024-09-11

    work page 2015

  3. [3]

    In: Proceedings of the 51st Annual International Symposium on Computer Architecture

    Wang, J., Berger, D.S., Kazhamiaka, F., Irvene, C., Zhang, C., Choukse, E., Frost, K., Fonseca, R., Warrier, B., Bansal, C., Stern, J., Bianchini, R., Sri- raman, A.: Designing Cloud Servers for Lower Carbon. In: Proceedings of the 51st Annual International Symposium on Computer Architecture. ISCA ’24, pp. 452–470. IEEE Press, Buenos Aires, Argentina (202...

    arXiv 2025

  4. [4]

    Technical report, GSMA (February 2025)

    Kamiya, G., Moore, S.: Rethinking mobile phones: The business case for circularity. Technical report, GSMA (February 2025). https://www.gsma. com/solutions-and-impact/connectivity-for-good/external-affairs/wp-content/ uploads/2025/04/Rethinking-Mobile-Phones Web.pdf

    work page 2025

  5. [5]

    Social Science Research Network, Rochester, NY (2024)

    Proserpio, D., Goli, A., Mangini, T., Lau, K., Yu, D.: The impact of sustainability programs on consumer purchase behavior: Evidence from Amazon. Social Science Research Network, Rochester, NY (2024). https://doi.org/10.2139/ssrn.5045830 . https://papers.ssrn.com/abstract=5045830 Accessed 2025-05-02

    work page doi:10.2139/ssrn.5045830 2024

  6. [6]

    In: Proceedings of the 49th Annual International Symposium on Computer Architecture

    Gupta, U., Elgamal, M., Hills, G., Wei, G.-Y., Lee, H.-H.S., Brooks, D., Wu, C.-J.: ACT: designing sustainable computer systems with an architectural car- bon modeling tool. In: Proceedings of the 49th Annual International Symposium on Computer Architecture. ISCA ’22, pp. 784–799. Association for Comput- ing Machinery, New York, NY, USA (2022). https://do...

    arXiv 2022

  7. [7]

    https: //www.dell.com/en-us/dt/corporate/social-impact/advancing-sustainability/ climate-action/product-carbon-footprints.htm

    Inc., D.T.: Product Carbon Footprint (PCF)|Dell USA (2024). https: //www.dell.com/en-us/dt/corporate/social-impact/advancing-sustainability/ climate-action/product-carbon-footprints.htm

    work page 2024

  8. [8]

    In: Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems

    Englhardt, Z., H¨ ahnlein, F., Mei, Y., Lin, T., Sun, C.M., Zhang, Z., Patel, S., Schulz, A., Iyer, V.: Incorporating Sustainability in Electronics Design: Obstacles and Opportunities. In: Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems. CHI ’25, pp. 1–17. Association for Computing Machin- ery, New York, NY, USA (2025). https:...

    arXiv 2025

Show all 64 references
  1. [9]

    Technical report, Asia-Pacific Economic Cooperation (February 2004)

    Lee, K.-M., Inaba, A.: Life Cycle Assessment Best Practices of ISO 14040 Series. Technical report, Asia-Pacific Economic Cooperation (February 2004)

    2004

  2. [10]

    One Earth4(12), 1704–1714 (2021) https://doi.org/10.1016/j.oneear.2021.11.007

    Bamana, G., Miller, J.D., Young, S.L., Dunn, J.B.: Addressing the social life cycle inventory analysis data gap: Insights from a case study of cobalt mining in the Democratic Republic of the Congo. One Earth4(12), 1704–1714 (2021) https://doi.org/10.1016/j.oneear.2021.11.007 ....

    2021 doi

  3. [11]

    Environmental Science & Technology58(44), 19595–19603 (2024) https://doi

    Tu, Q., Guo, J., Li, N., Qi, J., Xu, M.: Mitigating Grand Challenges in Life Cycle Inventory Modeling through the Applications of Large Language Models. Environmental Science & Technology58(44), 19595–19603 (2024) https://doi. org/10.1021/acs.est.4c07634 . Accessed 2025-06-28

    2024 doi

  4. [12]

    Journal of Industrial Ecology29(3), 955–966 (2025) https://doi.org/10.1111/jiec.70022

    Zhao, B., Jiang, J., Xu, M., Tu, Q.: A data-centric investigation on the challenges of machine learning methods for bridging life cycle inventory data gaps. Journal of Industrial Ecology29(3), 955–966 (2025) https://doi.org/10.1111/jiec.70022 . eprint: https://onlinelibrary.wi...

    2025 doi

  5. [13]

    Accessed 2025-01- 27

    Andrae, A.S.G., Andersen, O.: Life cycle assessments of consumer electronics — are they consistent? The International Journal of Life Cycle Assessment15(8), 827–836 (2010) https://doi.org/10.1007/s11367-010-0206-1 . Accessed 2025-01- 27

    2010 doi

  6. [14]

    Jinhui, , Singh, N.: Recycling of Spent Lithium-Ion Battery: A Critical Review

    Zeng, X., , L. Jinhui, , Singh, N.: Recycling of Spent Lithium-Ion Battery: A Critical Review. Critical Reviews in Environmental Science and Technol- ogy44(10), 1129–1165 (2014) https://doi.org/10.1080/10643389.2013.763578 . eprint: https://doi.org/10.1080/10643389.2013.763578...

    2014

  7. [15]

    National Academies Press, Washington, D.C

    Current Methods for Life Cycle Analyses of Low-Carbon Transportation Fuels in The United States. National Academies Press, Washington, D.C. (2022). https: //doi.org/10.17226/26402 . https://www.nap.edu/catalog/26402 Accessed 2025- 06-28

    2022 doi

  8. [16]

    Waste Management171, 281–291 (2023) https://doi.org/10.1016/j.wasman.2023.08.040

    Tan, Y., Wen, Z., Hu, Y., Zeng, X., Kosajan, V., Yin, G., Zhang, T.: Single-use plastic bag alternatives result in higher environmental impacts: Multi-regional analysis in country with uneven waste management. Waste Management171, 281–291 (2023) https://doi.org/10.1016/j.wasma...

    2023 doi

  9. [17]

    Balaji, B., Vunnava, V.S.G., Domingo, N., Gupta, S., Gupta, H., Guest, G., Srinivasan, A.: Flamingo: Environmental Impact Factor Matching for Life Cycle Assessment with Zero-shot Machine Learning. ACM J. Comput. Sustain. Soc. 1(2), 11–11123 (2023) https://doi.org/10.1145/36163...

    2023 doi

  10. [18]

    In: Proceedings of the 36th Annual 31 ACM Symposium on User Interface Software And Technology

    Lu, J., Desta, B., Wu, K.D., Nith, R., Passananti, J.E., Lopes, P.: ecoEDA: Recy- cling E-waste During Electronics Design. In: Proceedings of the 36th Annual 31 ACM Symposium on User Interface Software And Technology. UIST ’23, pp. 1–14. Association for Computing Machinery, Ne...

    2023

  11. [19]

    Zhang, Z., H¨ ahnlein, F., Mei, Y., Englhardt, Z., Patel, S., Schulz, A., Iyer, V.: DeltaLCA: Comparative Life-Cycle Assessment for Electronics Design. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.8(1), 29–12929 (2024) https: //doi.org/10.1145/3643561 . Accessed 2025-04-25

    2024 doi

  12. [20]

    https://www.fairphone

    IZM, F.: Life Cycle Assessment of the Fairphone 5 (2024). https://www.fairphone. com/wp-content/uploads/2024/06/Fairphone5 LCA Report 2024.pdf

    2024

  13. [21]

    OpenAI: GPT-4 Technical Report. arXiv. arXiv:2303.08774 [cs] (2023). https:// doi.org/10.48550/arXiv.2303.08774 . http://arxiv.org/abs/2303.08774 Accessed 2024-08-05

    2023 doi arXiv

  14. [22]

    al., G.T.G.: Gemini: A Family of Highly Capable Multimodal Models. arXiv. arXiv:2312.11805 [cs] (2025). https://doi.org/10.48550/arXiv.2312.11805 . http: //arxiv.org/abs/2312.11805 Accessed 2025-07-06

    2025 doi arXiv

  15. [23]

    Advances in Neural Information Processing Systems36, 34892–34916 (2023)

    Liu, H., Li, C., Wu, Q., Lee, Y.J.: Visual Instruction Tuning. Advances in Neural Information Processing Systems36, 34892–34916 (2023). Accessed 2025-07-06

    2023

  16. [24]

    Nature, 1–9 (2025) https://doi.org/10.1038/s41586-025-08866-7

    Tu, T., Schaekermann, M., Palepu, A., Saab, K., Freyberg, J., Tanno, R., Wang, A., Li, B., Amin, M., Cheng, Y., Vedadi, E., Tomasev, N., Azizi, S., Singhal, K., Hou, L., Webson, A., Kulkarni, K., Mahdavi, S.S., Semturs, C., Gottweis, J., Barral, J., Chou, K., Corrado, G.S., Ma...

    2025 doi

  17. [25]

    Bandiera abtest: a Cg type: News Feature Subject term: Machine learning, Lab life, Research management

    Jones, N.: What’s it like to work with an AI team of virtual scientists? Nature643(8070), 22–25 (2025) https://doi.org/10.1038/d41586-025-02028-5 . Bandiera abtest: a Cg type: News Feature Subject term: Machine learning, Lab life, Research management. Accessed 2025-07-06

    2025 doi

  18. [26]

    https://github.com/google/ travel-impact-model

    LLC, G.: Travel Impact Model 1.9.1 (2024). https://github.com/google/ travel-impact-model

    2024

  19. [27]

    https://www.gstatic.com/ gumdrop/sustainability/google-maps-eco-friendly-routing.pdf

    Google: Google Maps Eco-Friendly Routing (2021). https://www.gstatic.com/ gumdrop/sustainability/google-maps-eco-friendly-routing.pdf

    2021

  20. [28]

    https://apps.autodesk

    Autodesk: Manufacturing Sustainability Insights (2024). https://apps.autodesk. com/FUSION/en/Detail/Index?id=5902568957837856470&os=Mac&appLang= en

    2024

  21. [29]

    https://appsource

    Microsoft: Microsoft Sustainability Calculator (2024). https://appsource. 32 microsoft.com/en-us/product/power-bi/coi-sustainability.emissions impact dashboard

    2024

  22. [30]

    https://makersite.io/

    Makersite: Makersite (2024). https://makersite.io/

    2024

  23. [31]

    https://www.sluicebox.ai/

    Sluicebox: Sluicebox (2024). https://www.sluicebox.ai/

    2024

  24. [32]

    https://netzero.imec-int.com/

    Centre, I.M.: imec.netzero (2024). https://netzero.imec-int.com/

    2024

  25. [33]

    https://www.2030calculator.com/

    AB, O.S.: The 2030 Calculator (2023). https://www.2030calculator.com/

    2023

  26. [34]

    Chatty, T., Moeller, B.L., Pantelimon, I.A., Parnell, C.D., Khan, T.M., Laurin, L., Faludi, J., Murnane, E.L.: EcoSketch: Promoting Sustainable Design through Iterative Environmental Assessment during Early-Stage Product Development. ACM J. Comput. Sustain. Soc.2(2), 22–12229 ...

    2024

  27. [35]

    Ma, Z., Huang, W., Zhang, J., Gupta, T., Krishna, R.: m&m’s: A Benchmark to Evaluate Tool-Use for multi-step multi-modal Tasks. (2024). https://doi.org/10. 48550/arXiv.2403.11085 . arXiv:2403.11085 [cs].http://arxiv.org/abs/2403.11085 Accessed 2025-06-24

    2024 doi arXiv

  28. [36]

    https://www.delltechnologies.com/asset/en-us/products/ multi-product/industry-market/pcf-lca-whitepaper.pdf

    Dell Inc: Understanding Life Cycle Assessments (LCAs), Product Carbon Foot- prints (PCFs), and the uses and limitations of PAIA, a streamlined LCA methodology (2023). https://www.delltechnologies.com/asset/en-us/products/ multi-product/industry-market/pcf-lca-whitepaper.pdf

    2023

  29. [37]

    The International Journal of Life Cycle Assessment21(9), 1218–1230 (2016) https: //doi.org/10.1007/s11367-016-1087-8

    Wernet, G., Bauer, C., Steubing, B., Reinhard, J., Moreno-Ruiz, E., Weidema, B.: The ecoinvent database version 3 (part I): overview and methodology. The International Journal of Life Cycle Assessment21(9), 1218–1230 (2016) https: //doi.org/10.1007/s11367-016-1087-8 . Accessed...

    2016 doi

  30. [38]

    https://www.fairphone

    IZM, F.: Life Cycle Assessment of the Fairphone 4 (2022). https://www.fairphone. com/wp-content/uploads/2022/07/Fairphone-4-Life-Cycle-Assessment-22.pdf

    2022

  31. [39]

    Zhang, Z., Thavikulwat, P., Metzger, A.L., Mei, Y., H¨ ahnlein, F., Englhardt, Z., Abowd, G.D., Patel, S., Schulz, A., Cheng, T., Iyer, V.: Living Sustainability: In- Context Interactive Environmental Impact Communication. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.9...

    2025

  32. [40]

    Microsoft: Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone. arXiv. arXiv:2404.14219 [cs] (2024). https://doi.org/10.48550/ arXiv.2404.14219 . http://arxiv.org/abs/2404.14219 Accessed 2025-07-06

    2024 doi arXiv

  33. [41]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Chen, B., Xu, Z., Kirmani, S., Ichter, B., Sadigh, D., Guibas, L., Xia, F.: 33 SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capa- bilities. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14455–14465 (2024). http...

    2024 doi arXiv

  34. [42]

    ecoinvent: ecoinvent - Data with purpose. (2024). https://ecoinvent.org/ Accessed 2024-09-02

    2024

  35. [43]

    In: Extended Abstracts of the 2022 CHI Conference on Human Factors in Computing Systems

    Arroyos, V., Viitaniemi, M.L.K., Keehn, N., Oruganti, V., Saunders, W., Strauss, K., Iyer, V., Nguyen, B.H.: A Tale of Two Mice: Sustainable Electronics Design and Prototyping. In: Extended Abstracts of the 2022 CHI Conference on Human Factors in Computing Systems. CHI EA ’22,...

    2022

  36. [44]

    In: Proceedings of the ACM Web Conference 2023, pp

    Balaji, B., Vunnava, V.S.G., Guest, G., Kramer, J.: CaML: Carbon Foot- printing of Household Products with Zero-Shot Semantic Text Similar- ity. In: Proceedings of the ACM Web Conference 2023, pp. 4004–4014. ACM, Austin TX USA (2023). https://doi.org/10.1145/3543507.3583882 . ...

    2023

  37. [45]

    Environmental Science & Technology59(18), 9113–9122 (2025) https://doi.org/10.1021/acs.est.4c12667

    Balaji, B., Ebrahimi, F., G Domingo, N.G., Vunnava, V.S.G., Faridee, A.-Z., Ramalingam, S., Gupta, S., Wang, A., Gupta, H., Belcastro, D., Axten, K., Hakian, J., Kramer, J., Srinivasan, A., Tu, Q.: Emission Factor Recommenda- tion for Life Cycle Assessments with Generative AI....

    2025 doi

  38. [46]

    https://www

    Maps, E.: Electricity Maps Carbon Intensity Data (2025). https://www. electricitymaps.com

    2025

  39. [47]

    Journal of Chiropractic Medicine15(2), 155–163 (2016) https://doi.org/10.1016/j.jcm.2016.02.012

    Koo, T.K., Li, M.Y.: A Guideline of Selecting and Reporting Intraclass Correla- tion Coefficients for Reliability Research. Journal of Chiropractic Medicine15(2), 155–163 (2016) https://doi.org/10.1016/j.jcm.2016.02.012 . Accessed 2026-01-06

    2016 doi

  40. [48]

    Proceedings of the National Academy of Sciences121(27), 2311878121 (2024) https://doi.org/10.1073/pnas.2311878121

    Bahri, Y., Dyer, E., Kaplan, J., Lee, J., Sharma, U.: Explaining neural scal- ing laws. Proceedings of the National Academy of Sciences121(27), 2311878121 (2024) https://doi.org/10.1073/pnas.2311878121 . Accessed 2025-06-24

    2024 doi

  41. [49]

    Kaplan, J., McCandlish, S., Henighan, T., Brown, T.B., Chess, B., Child, R., Gray, S., Radford, A., Wu, J., Amodei, D.: Scaling Laws for Neural Language Models. arXiv. arXiv:2001.08361 [cs] (2020). https://doi.org/10.48550/arXiv. 2001.08361 . http://arxiv.org/abs/2001.08361 Ac...

    2020 doi arXiv

  42. [50]

    Python library

    OpenAI: OpenAI Agents SDK. Python library. Version v0.0.19 (latest release as of June 18, 2025) (2025). https://github.com/openai/openai-agents-python 34

    2025

  43. [51]

    Nature Computational Science5(10), 962–972 (2025) https://doi.org/10.1038/ s43588-025-00849-y

    Ding, K., Yu, J., Huang, J., Yang, Y., Zhang, Q., Chen, H.: SciToolA- gent: a knowledge-graph-driven scientific agent for multitool integration. Nature Computational Science5(10), 962–972 (2025) https://doi.org/10.1038/ s43588-025-00849-y . Accessed 2026-01-17

    2025

  44. [52]

    https://www.ifixit.com/

    iFixit: iFixit (2024). https://www.ifixit.com/

    2024

  45. [53]

    https://www.techpowerup.com/

    TechPowerUp: TechPowerUp (2024). https://www.techpowerup.com/

    2024

  46. [54]

    https://laptopmedia.com/

    LaptopMedia: LaptopMedia (2024). https://laptopmedia.com/

    2024

  47. [55]

    In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: You Only Look Once: Uni- fied, Real-Time Object Detection. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779–788 (2016). https://doi.org/ 10.1109/CVPR.2016.91 .https://ieeexplore.ieee.org/doc...

    2016

  48. [56]

    (2020).https://openreview.net/forum?id=YicbFdNTTyAccessed 2024-11-29

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., Houlsby, N.: An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. (2020).https://openreview.net/for...

    2020

  49. [57]

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., Lo, W.-Y., Doll´ ar, P., Girshick, R.: Segment Anything. arXiv. arXiv:2304.02643 [cs] (2023). https://doi.org/10.48550/arXiv.2304.02643 . http://arxiv.org/abs/2304.026...

    2023 doi arXiv

  50. [58]

    Energy & Environmental Science7(11), 3674–3682 (2014) https://doi.org/10.1039/C4EE01995D

    Liu, J., Yang, C., Wu, H., Lin, Z., Zhang, Z., Wang, R., Li, B., Kang, F., Shi, L., Wong, C.P.: Future paper based printed circuit boards for green electronics: fabrication and life cycle assessment. Energy & Environmental Science7(11), 3674–3682 (2014) https://doi.org/10.1039...

    2014 doi

  51. [59]

    In: Proceedings of the 2009 International Conference on Computer- Aided Design

    Zolotov, V., Visweswariah, C., Xiong, J.: Voltage binning under process vari- ation. In: Proceedings of the 2009 International Conference on Computer- Aided Design. ICCAD ’09, pp. 425–432. Association for Computing Machin- ery, New York, NY, USA (2009). https://doi.org/10.1145...

    2009

  52. [60]

    https://www.energystar.gov/

    Star, E.: Energy Star (2024). https://www.energystar.gov/

    2024

  53. [61]

    https: //www.epeat.net/

    Registry, E.: Electronic Product Environmental Assessment Tool (2024). https: //www.epeat.net/

    2024

  54. [62]

    https://tcocertified.com/

    Professional Employees, S.C.: TCO Certified (2024). https://tcocertified.com/

    2024

  55. [63]

    https: //sustainability.aboutamazon.com/devices-carbon-methodology.pdf 35

    Amazon: Amazon Devices Product Carbon Footprint Methodology (2023). https: //sustainability.aboutamazon.com/devices-carbon-methodology.pdf 35

    2023

  56. [64]

    https://www.delltechnologies.com/asset/en-us/products/laptops-and-2-in-1s/ technical-support/dell-xps-17-9730-pcf-datasheet.pdf 36

    Inc., D.T.: Product Carbon Footprint Dell XPS 17 9730 (2023). https://www.delltechnologies.com/asset/en-us/products/laptops-and-2-in-1s/ technical-support/dell-xps-17-9730-pcf-datasheet.pdf 36

    2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.