Pith. sign in

REVIEW 3 major objections 6 minor 64 references

AI4EF: Artificial Intelligence for Energy Efficiency in the Building Sector

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read AI4EF claims its ML services turn a few building parameters into reliable retrofit and PV recommendations.

desk verdict A well-described open-source building retrofit tool whose stated real-world validation is asserted without numbers; the engineering is solid, the evidence is not. read the letter →

arxiv 2412.04045 v1 pith:EBMNZGB2 submitted 2024-12-05 cs.LG

classification cs.LG
keywords energyefficiencymachinelearningbuildingretrofittingphotovoltaicinstallationdecisionsupportpredictivemodelingMLOpsEuropeandataspace
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

AI4EF is a software platform for building energy decision-making. The paper's central claim is that its machine-learning services can turn a few building parameters into tailored recommendations for retrofits and photovoltaic installations, estimating energy savings, costs, payback periods, and CO2 reductions. A training environment lets data scientists replace the default models with custom ones, and integration with a European energy data space provides additional data for training. The authors assert the tool was built on real data and used by actual stakeholders, although the manuscript itself reports no user study or accuracy numbers.

What carries the argument

The load-bearing mechanism is a modular microservices backend with two MLApp sub-services, Building Retrofitting and Photovoltaic Installation, plus a Training Playground orchestrated by a workflow-management tool. The Training Playground runs an Ingestion-Training-Evaluation pipeline: it retrieves and cleans data, trains a multilayer perceptron with a deep-learning training library while a hyperparameter optimizer searches configurations, and evaluates predictions with regression or classification metrics. This pipeline is what gives AI4EF its claimed flexibility: the default models can be retrained on user-provided data, and the versioned model checkpoints are what the two dashboards call to return recommendations. An open-source identity-management layer and a data-space connector support the platform's claimed suitability for secure, multi-organization deployment.

What would settle it

Inspect the released version 0.0.2 and rerun the Training Playground pipeline on a held-out set of building records with known post-retrofit outcomes; if predicted energy savings, payback periods, or CO2 reductions are not compared against measured results or a simple baseline and show large errors, the platform's central promise of reliable, user-satisfying recommendations is false.

Watch

Extended reading notes

Core claim

The paper claims that AI4EF closes a gap left by both open-source and proprietary building-energy tools. Its Building Retrofitting service takes inputs such as building total area, number of floors, current energy class, and energy consumption, and returns recommended construction works, expected primary-energy reduction, and CO2 savings; its Photovoltaic Installation service takes consumption, electricity price, and equipment costs and predicts generation, self-consumption, financial savings, and payback period. Unlike tools that require full building geometry and HVAC details, AI4EF is designed to produce initial insights from minimal data. Its Training Playground wraps data ingestion, hyperparameter optimization, training, and evaluation in an MLOps pipeline so users can retrain the underlying multilayer-perceptron models on their own datasets. The paper further claims the platform was developed using real data based on the needs of a Latvian environmental fund and was evaluated and used by real users satisfying their needs.

Load-bearing premise

The load-bearing premise is that AI4EF was actually built on real data and evaluated by real users whose needs it satisfied, because the paper's stated contribution over its competitors rests on that validation; the manuscript offers no user study, dataset, or metrics to back it up.

Editorial extensions

If this is right

  • Building owners and managers can obtain retrofit and photovoltaic recommendations from a small set of inputs rather than full building geometry and HVAC specifications.
  • Data scientists can retrain the underlying models on their own datasets through the Training Playground, making recommendations specific to a building stock or region.
  • Government representatives and energy consultants can use the platform to support evidence-based policy and advisory decisions on energy-efficiency investments.
  • Connection to a European energy data space allows the platform to draw on shared, large-scale datasets for training and analysis.
  • The open-source, modular architecture makes the platform adaptable to different regulatory and organizational contexts.

Reading between the lines

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

  • Editorial extension: if the released models were run against measured post-retrofit consumption data, the platform's minimal-data advantage could be quantified; the paper does not supply that benchmark.
  • Editorial extension: the data-space connector implies a path toward continuously retrained models fed by live meter streams, turning the Training Playground into a monitoring-and-updating loop rather than a one-off training tool; the paper presents this only implicitly as future work.
  • Editorial extension: the same input-parameter-to-recommendation pattern could be lifted to other investment decisions such as heat-pump sizing or district-heating connections, since the ML core is not tied to retrofit-specific features; the paper does not claim this.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. AI4EF is presented as a modular, AI-based decision-support platform for building energy retrofitting and photovoltaic installation assessment. The paper describes the system architecture (Docker microservices, React frontend, Python/MLOps backend), the Training Playground for custom model training, and integration with the Enershare Data Space. The introduction claims that AI4EF is open-source, customizable, validated with real users at the Latvian Environmental Investment Fund (LEIF), and integrated with European energy data sharing. The body includes illustrative UI screenshots and example input tables but no quantitative evaluation of the platform's predictive models or user satisfaction.

Significance. The paper's main contribution is the architectural and software engineering description of AI4EF, which combines building retrofit recommendations, PV assessment, and an MLOps training environment in one platform. If the claimed real-world validation were documented with data and metrics, this would be a useful addition to the building energy efficiency software landscape, especially for integrating European Data Space standards. The code repository and use of standard ML tools are commendable. However, the current version lacks the evidence needed to substantiate the central claims of reliable predictions and user satisfaction, limiting the paper's immediate impact.

major comments (3)
  1. [Section 1 (third bullet), Section 4] The third contribution bullet in Section 1 and the Impact section state that AI4EF 'was built on real data based on the needs of LEIF' and 'was evaluated and utilised by real users satisfying their needs.' However, the manuscript provides no dataset description, no sample size, no evaluation protocol, no prediction accuracy or satisfaction measures, and no user study details. This is a load-bearing claim: without such evidence, the assertion that AI4EF meets real-world demands and offers reliable recommendations is unsupported. Please either add a validation section with quantitative results (e.g., model performance on a held-out LEIF dataset, user surveys, or case studies) or revise the claims to reflect the current evidence level.
  2. [Section 3, Tables 2 and 3, Figures 3, 5, 6] The illustrative examples in Section 3 show only screenshots and input parameter tables; no quantitative results from the Building Retrofitting or Photovoltaic models are reported. The Training Playground is said to produce metrics and visualizations (Figures 5 and 6), but no metrics from the actual LEIF deployment or any benchmark are given. Without prediction accuracy (e.g., MAE, RMSE, classification accuracy, F1), error bars, or a comparison to existing tools (OpenStudio, Aurora Solar, etc.), the central claim that AI4EF provides reliable 'tailored recommendations' is not established. Please include an evaluation section with concrete numbers and, ideally, ablations or baseline comparisons.
  3. [Section 1 (first bullet), Table 1] The first contribution bullet describes 'Open-Source Accessibility' and contrasts AI4EF with proprietary tools, but Table 1 reports the legal license as 'Attribution-Noncommercial 4.0 International' (CC BY-NC). This license restricts commercial use and is not an OSI-approved open-source license, which conflicts with the paper's open-source claim. Please clarify the licensing terms and align the description with the actual license, or reconsider the claim.
minor comments (6)
  1. [Section 2.2] The heading 'Software functionalities' is immediately followed by '2.3 Interface Features' with no content; please either provide the intended discussion or remove the empty subsection.
  2. [Nomenclature] 'HV AC' should be 'HVAC' (the term is used inconsistently in the text).
  3. [References] Reference [12] contains a garbled author list; please fix the citation.
  4. [Section 3.2, Figures] Figure numbering appears out of order: Section 3.2 refers to 'Figure 7' before Figure 8, and some screenshots are only in a later appendix; consider renumbering for clarity.
  5. [Table 1, Section 1] Table 1 lists the code and software licenses as 'Attribution-Noncommercial 4.0 International', but the text uses 'open-source' in the introduction; please use consistent terminology.
  6. [Section 3.1, Table 3] In Section 3.1, the text says users can leave the energy generation field blank in the PV sub-service, but the example input in Table 3 shows a placeholder '-' for that field; verify consistency.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an architecture and system description, not a derivation, and no prediction is shown to reduce to a fitted input or self-citation.

full rationale

AI4EF is a software platform description, not a derivation chain. The paper contains no equations connecting outputs to inputs by construction, and no model is fitted to a subset of data and then presented as predicting that same data. The closest candidate is the Section 1 claim that the tool 'was built on real data based on the needs of LEIF, and was evaluated and utilised by real users satisfying their needs.' That is an unsupported validation assertion, but it is an evidence gap (no dataset description, no accuracy or satisfaction metrics, no user study), not a circular reduction: it does not define a quantity in terms of itself, nor does it reuse a fitted parameter as a prediction. The only self-citation, reference [1], points to the project's code repository and release metadata, which is normal software-paper practice and is not load-bearing for any scientific claim. The Training Playground generates model metrics, but the paper reports no specific predictions derived from those metrics, so there is no fitted-input-called-prediction step to exhibit. Under the rule that circularity must be demonstrated with a specific quote and reduction, no circular step can be identified here; the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

The paper reports no fitted parameter values, and the central premises are the unvalidated effectiveness of the ML models and the unverified user validation claim. No new physical or conceptual entities are introduced.

assumptions (2)
  • domain assumption The machine learning models can produce accurate and useful predictions of retrofit outcomes from the small set of input parameters listed in Table 2.
    The platform's value depends entirely on model accuracy, but the paper provides no metrics, error analysis, or validation results to support this assumption. It is implicitly invoked in Sections 1 and 3.
  • ad hoc to paper The tool was evaluated and utilized by real users satisfying their needs, as stated in Section 1, bullet 3.
    This is a direct assertion in the text with no supporting evidence. It is the basis for the 'Real-World Validation' contribution claim.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AI4EF: Artificial Intelligence for Energy Efficiency in the Building Sector." pith.science (2026). https://pith.science/paper/EBMNZGB2

@misc{pith2026241204045,
  author       = {Pith},
  title        = {Pith review of: AI4EF: Artificial Intelligence for Energy Efficiency in the Building Sector},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EBMNZGB2}},
  note         = {Machine review of arXiv:2412.04045}
}
read the original abstract

AI4EF, Artificial Intelligence for Energy Efficiency, is an advanced, user-centric tool designed to support decision-making in building energy retrofitting and efficiency optimization. Leveraging machine learning (ML) and data-driven insights, AI4EF enables stakeholders such as public sector representatives, energy consultants, and building owners to model, analyze, and predict energy consumption, retrofit costs, and environmental impacts of building upgrades. Featuring a modular framework, AI4EF includes customizable building retrofitting, photovoltaic installation assessment, and predictive modeling tools that allow users to input building parameters and receive tailored recommendations for achieving energy savings and carbon reduction goals. Additionally, the platform incorporates a Training Playground for data scientists to refine ML models used by said framework. Finally, AI4EF provides access to the Enershare Data Space to facilitate seamless data sharing and access within the ecosystem. Its compatibility with open-source identity management, Keycloak, enhances security and accessibility, making it adaptable for various regulatory and organizational contexts. This paper presents an architectural overview of AI4EF, its application in energy efficiency scenarios, and its potential for advancing sustainable energy practices through artificial intelligence (AI).

Figures

Figures reproduced from arXiv: 2412.04045 by the authors.

Figure 1
Figure 1. AI4EF - Data space Integration 2.2. Software functionalities 2.3. Interface Features AI4EF can be accessed and be navigated through its main UI, which serves as a centralized dashboard for all sub-services of the tool. Homepage:. Figure 3b shows the homepage displayed after signing in. Here, an overview of the primary tools is presented in a visually organized layout, with clear headings that introduce the key servi… view at source ↗
Figure 2
Figure 2. AI4EF - Software Architecture Diagram Ingestion. This step is separated into two distinct categories: the helper functions and the main assets. Starting from the helper functions, these include methods for validating file paths, URLs, and database connection strings, as well as for retrieving data from various sources like databases, local files, or APIs. The main assets handle data extraction, cleaning, and scaling… view at source ↗
Figure 3
Figure 3. AI4EF - MLApp Dashboard tation. Evaluation. This step sets up an evaluation pipeline to test model perfor￾mance, generate metrics, and visualize HPO results. It computes different evaluation metrics depending on the sub-service: regression or classification. Visual summaries of the HPO process show parameter importance and per￾formance trends. All metrics and plots are stored in a structured format, providing a clea… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: AI4EF - Training Playground Sequence Diagram [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: AI4EF - Training Playground - Ingestion and Training Pipelines [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: AI4EF - Evaluation pipeline energy sources and retrofitting projects, thereby fostering the transition to￾ward more sustainable building practices. The modular design of the plat￾form allows for continuous updates and improvements, ensuring that users benefit from the …
Figure 7
Figure 7. Figure 7: AI4EF - Training Playground - ML Pipeline [PITH_FULL_IMAGE:figures/full_fig_p021_7.png]
Figure 8
Figure 8. Figure 8: AI4EF - Excel sheet used to calculate energy savings and greenhouse gas emis [PITH_FULL_IMAGE:figures/full_fig_p022_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 64 canonical work pages

  1. [1]

    epu-ntua/enershare-ai4ef: 0.0.2, November 2024

    Alexandros Tzortzis and Sotiris Pelekis. epu-ntua/enershare-ai4ef: 0.0.2, November 2024. Creative Commons Attribution 4.0 International

  2. [2]

    Scott Chacon and Ben Straub. Pro git. Apress, 2014

  3. [3]

    Python reference manual

    Guido Van Rossum and Fred L Drake Jr. Python reference manual . Centrum voor Wiskunde en Informatica Amsterdam, 1995

  4. [4]

    Python package index - pypi, 2024

    Python Software Foundation. Python package index - pypi, 2024. Ac- cessed: 04-10-2024

  5. [5]

    Sustainable development goals, n.d

    European Union. Sustainable development goals, n.d. Accessed: [Date of access]

  6. [6]

    European Commission, Directorate-General for Energy. Non paper on complementary economic modelling undertaken by dg ener analysing the impacts of overall energy efficiency target of 13% to 19% in the context of discussions in the european parliament on the revision of the energy efficiency directive, June 2022

  7. [7]

    Repowereu: Affordable, secure and sustainable energy for europe, 2024

    European Commission. Repowereu: Affordable, secure and sustainable energy for europe, 2024. Accessed: 2024-11-07

  8. [8]

    A review on sustainable design of renewable energy systems

    Long Shi and Michael Yit Lin Chew. A review on sustainable design of renewable energy systems. Renewable and Sustainable Energy Reviews , 16:192–207, 1 2012

Show all 64 references
  1. [9]

    Renewable energy and the city: Urban life in an age of fossil fuel depletion and climate change

    Peter Droege. Renewable energy and the city: Urban life in an age of fossil fuel depletion and climate change. Bulletin of Science, Technology & Society, 22(2):87–99, 2002

  2. [10]

    Holechek, Hatim M

    Jerry L. Holechek, Hatim M. E. Geli, Mohammed N. Sawalhah, and Raul Valdez. A global assessment: Can renewable energy replace fossil fuels by 2050? Sustainability, 14(8), 2022

  3. [11]

    Role of energy storage systems in energy transition from fossil fuels to renewables

    Anam Kalair, Naeem Abas, Muhammad Shoaib Saleem, Ali Raza Kalair, and Nasrullah Khan. Role of energy storage systems in energy transition from fossil fuels to renewables. Energy Storage, 3(1), February 2020

  4. [12]

    Research and case appli- cation of zero-carbon buildings based on multi-system integration func- tion

    J ; Zhang, Y ; Lin, L ; Wang, Q ; Wang, K ; Zhu, S ; Yang, G Guo, Maria Mauro, Shi-Jie Cao, Jiaji Zhang, Yuting Lin, Lan Wang, Qiankun 15 Wang, Ke Zhu, Shize Yang, and Guoqing Guo. Research and case appli- cation of zero-carbon buildings based on multi-system integration func-...

  5. [13]

    The challenge of sus- tainable building renovation: assessment of current criteria and future outlook

    Olatz Pombo, Beatriz Rivela, and Javier Neila. The challenge of sus- tainable building renovation: assessment of current criteria and future outlook. Journal of Cleaner Production , 123:88–100, 6 2016

  6. [14]

    Systematic method for the sustainability analysis of refurbishment concepts of exterior walls

    Tarja H¨ akkinen. Systematic method for the sustainability analysis of refurbishment concepts of exterior walls. Construction and Building Materials, 37:783–790, 12 2012

  7. [15]

    Sustainabil- ity of heating, ventilation and air-conditioning (hvac) systems in build- ings—an overview

    Nilofar Asim, Marzieh Badiei, Masita Mohammad, Halim Razali, Armin Rajabi, Lim Chin Haw, and Mariyam Jameelah Ghazali. Sustainabil- ity of heating, ventilation and air-conditioning (hvac) systems in build- ings—an overview. International Journal of Environmental Research and P...

  8. [16]

    Optimization of energy in sustainable architecture and green roofs in construction: A review of challenges and advantages

    Sara ziaee, Zeynab Gholampour, Mina Soleymani, Parisa Doraj, Omid Hossein Eskandani, and Samireh Kadaei. Optimization of energy in sustainable architecture and green roofs in construction: A review of challenges and advantages. Complexity, 2022(1), January 2022

  9. [17]

    Green archi- tecture: A concept of sustainability

    Amany Ragheb, Hisham El-Shimy, and Ghada Ragheb. Green archi- tecture: A concept of sustainability. Procedia - Social and Behavioral Sciences, 216:778–787, 2016. Urban Planning and Architectural Design for Sustainable Development (UPADSD)

  10. [18]

    Building energy- efficiency standards in a life cycle primary energy perspective

    Ambrose Dodoo, Leif Gustavsson, and Roger Sathre. Building energy- efficiency standards in a life cycle primary energy perspective. Energy and Buildings , 43(7):1589–1597, 2011

  11. [19]

    The role of green building ma- terials in sustainable architecture: Innovations, challenges, and future trends

    Obinna Iwuanyanwu, Ifechukwu Gil-Ozoudeh, Azubuike Chukwudi Ok- wandu, and Chidiebere Somadina Ike. The role of green building ma- terials in sustainable architecture: Innovations, challenges, and future trends. International Journal of Applied Research in Social Sciences , 6(...

  12. [20]

    Smart materials innovative technologies in architecture; towards innovative design paradigm

    Abeer Samy Yousef Mohamed. Smart materials innovative technologies in architecture; towards innovative design paradigm. Energy Procedia, 115:139–154, 2017. International Conference – Alternative and Renew- able Energy Quest, AREQ 2017, 1-3 February 2017, Spain. 16

  13. [21]

    A comprehensive review on emerging trends in smart green building technologies and sus- tainable materials

    Monika Patil, Samrudhi Boraste, and Pravin Minde. A comprehensive review on emerging trends in smart green building technologies and sus- tainable materials. Materials Today: Proceedings, 65:1813–1822, 2022. International Conference on Advances in Construction Materials and Structures

  14. [22]

    Review of energy- efficient hvac technologies for sustainable buildings

    Tosin Michael Olatunde, Azubuike Chukwudi Okwandu, Dorcas Oluwa- juwonlo Akande, and Zamathula Queen Sikhakhane. Review of energy- efficient hvac technologies for sustainable buildings. International Jour- nal of Science and Technology Research Archive , 6(2):012–020, April 2024

  15. [23]

    Green smart building: Requisites, archi- tecture, challenges, and use cases

    Pijush Kanti Dutta Pramanik, Bulbul Mukherjee, Saurabh Pal, Tanmoy Pal, and Simar Preet Singh. Green smart building: Requisites, archi- tecture, challenges, and use cases. In Green building management and smart automation , pages 1–50. IGI Global, 2020

  16. [24]

    Use of smart technologies for energy efficiency, energy- and load management in small and medium sized enterprises (smes)

    Lukas Hilger, Thorsten Schneiders, Frederic Pascal Meyer, and Jean Pierre Kroll. Use of smart technologies for energy efficiency, energy- and load management in small and medium sized enterprises (smes). 2018 7th International Energy and Sustainability Conference, IESC 2018 , 8 2018

  17. [25]

    Environmental impact of renewable energy source based electrical power plants: Solar, wind, hydroelectric, biomass, geothermal, tidal, ocean, and osmotic

    Abidur Rahman, Omar Farrok, and Md Mejbaul Haque. Environmental impact of renewable energy source based electrical power plants: Solar, wind, hydroelectric, biomass, geothermal, tidal, ocean, and osmotic. Renewable and Sustainable Energy Reviews , 161:112279, 2022

  18. [26]

    Har- nessing machine learning for sustainable futures: advancements in re- newable energy and climate change mitigation

    Kingsley Ukoba, Oluwatayo Racheal Onisuru, and Tien-Chien Jen. Har- nessing machine learning for sustainable futures: advancements in re- newable energy and climate change mitigation. Bulletin of the National Research Centre 2024 48:1 , 48:1–15, 10 2024

  19. [27]

    How does artificial intelligence affect pollutant emissions by improving energy efficiency and developing green technology

    Wei Zhou, Yan Zhuang, and Yan Chen. How does artificial intelligence affect pollutant emissions by improving energy efficiency and developing green technology. Energy Economics, 131:107355, 2024

  20. [28]

    Method and evaluations of the effective gain of artificial intelligence models for reduc- ing co2 emissions

    Paul Delano¨ e, Dieudonn´ e Tchuente, and Guillaume Colin. Method and evaluations of the effective gain of artificial intelligence models for reduc- ing co2 emissions. Journal of Environmental Management , 331:117261, 2023

  21. [29]

    Machine learning for optimising 17 renewable energy and grid efficiency

    Domenico Toscano, Grazia Fattoruso, Bankole I Oladapo, Mattew A Olawumi, and Francis T Omigbodun. Machine learning for optimising 17 renewable energy and grid efficiency. Atmosphere 2024, Vol. 15, Page 1250, 15:1250, 10 2024

  22. [30]

    Hooman Farzaneh, Ladan Malehmirchegini, Adrian Bejan, Taofeek Afo- labi, Alphonce Mulumba, and Precious P. Daka. Artificial intelligence evolution in smart buildings for energy efficiency. Applied Sciences , 11(2):763, January 2021

  23. [31]

    Navigating the nexus of artificial intelligence and renewable energy for the advancement of sustainable development goals

    Raghu Raman, Sangeetha Gunasekar, Deepa Kaliyaperumal, and Prema Nedungadi. Navigating the nexus of artificial intelligence and renewable energy for the advancement of sustainable development goals. Sustain- ability 2024, Vol. 16, Page 9144 , 16:9144, 10 2024

  24. [32]

    A review of the applications of artificial intelligence and big data to buildings for energy-efficiency and a com- fortable indoor living environment

    Muhammad Uzair Mehmood, Daye Chun, Zeeshan, Hyunjoo Han, Gyuyeob Jeon, and Kuan Chen. A review of the applications of artificial intelligence and big data to buildings for energy-efficiency and a com- fortable indoor living environment. Energy and Buildings , 202:109383, 2019

  25. [33]

    Utility grid: Present challenges and their potential solutions

    Eklas Hossain, Jakir Hossain, and Fuad Un-Noor. Utility grid: Present challenges and their potential solutions. IEEE Access, 6:60294–60317, 2018

  26. [34]

    Massoud Amin

    S. Massoud Amin. Smart grid: Overview, issues and opportunities. advances and challenges in sensing, modeling, simulation, optimization and control. European Journal of Control , 17(5):547–567, 2011

  27. [35]

    Towards net- zero emissions in the eu energy system by 2050

    Tsiropoulos I, Nijs W, Tarvydas D, and Ruiz Castello P. Towards net- zero emissions in the eu energy system by 2050. Technical Report KJ- NA-29981-EN-N (online),KJ-NA-29981-EN-C (print), Publications Of- fice of the European Union, Luxembourg, 2020

  28. [36]

    equest, 2023

    Electric Power Research Institute and Lawrence Berkeley National Lab- oratory. equest, 2023. Proprietary software for building energy simula- tion. Not open source

  29. [37]

    BuildingIQ

    Inc. BuildingIQ. Buildingiq, 2023. Proprietary energy management software for buildings

  30. [38]

    Autodesk

    Inc. Autodesk. Autodesk revit with dynamo, 2023. Proprietary BIM software with a visual programming tool (Dynamo) for customization

  31. [39]

    Aurora Solar

    Inc. Aurora Solar. Aurora solar, 2023. Proprietary photovoltaic system design and simulation software. 18

  32. [40]

    Openstudio, 2023

    National Renewable Energy Laboratory. Openstudio, 2023. Open-source software for building energy modeling and analysis

  33. [41]

    Retrofit advisor, 2023

    Retrofit Advisor. Retrofit advisor, 2023. Proprietary tool for retrofit recommendations in buildings

  34. [42]

    Folsom Labs

    Inc. Folsom Labs. Helioscope, 2023. Proprietary software for PV system design and simulation

  35. [43]

    Latvian environmental invest- ment fund (leif), 2024

    Latvian Environmental Investment Fund. Latvian environmental invest- ment fund (leif), 2024. Focused on sustainable environmental investment and development projects in Latvia

  36. [44]

    Enershare — the energy data space for europe

    Enershare — The Energy Data Space for Europe. Enershare — the energy data space for europe. https://enershare.eu/, 2024

  37. [45]

    Solar-log, 2023

    Solar-Log GmbH. Solar-log, 2023. Proprietary software for photovoltaic monitoring and analytics

  38. [46]

    Valentin Software

    Inc. Valentin Software. Pv*sol, 2023. Proprietary photovoltaic design and simulation software

  39. [47]

    Docker: Accelerated, containerized application development

    Docker. Docker: Accelerated, containerized application development. https://www.docker.com/, 2024. Accessed: 04-10-2024

  40. [48]

    Keycloak: Open source identity and access management,

    Keycloak. Keycloak: Open source identity and access management,

  41. [49]

    React: A javascript library for building user interfaces, 2024

    React. React: A javascript library for building user interfaces, 2024. Accessed: 04-10-2024

  42. [50]

    HTML5 Specification

    World Wide Web Consortium (W3C). HTML5 Specification. World Wide Web Consortium (W3C), 2014. Accessed: 2024-11-07

  43. [51]

    Cascading Style Sheets (CSS) Specification

    World Wide Web Consortium (W3C). Cascading Style Sheets (CSS) Specification. World Wide Web Consortium (W3C), 2018. Accessed: 2024-11-07

  44. [52]

    ECMAScript 2015 Language Specification

    ECMA International. ECMAScript 2015 Language Specification. ECMA International, 2015. Accessed: 2024-11-07

  45. [53]

    MUI: Material-UI React Component Library

    MUI Team. MUI: Material-UI React Component Library . MUI, 2024. Accessed: 2024-11-07

  46. [54]

    Dagster: The data orchestrator for machine learning, analytics, and etl, 2024

    Dagster. Dagster: The data orchestrator for machine learning, analytics, and etl, 2024. Accessed: 04-10-2024. 19

  47. [55]

    Scikit-learn: Machine learning in python, 2024

    Scikit-learn. Scikit-learn: Machine learning in python, 2024. Accessed: 04-10-2024

  48. [56]

    Scipy: Open source scientific tools for python, 2024

    SciPy. Scipy: Open source scientific tools for python, 2024. Accessed: 04-10-2024

  49. [57]

    Pytorch lightning: The lightweight pytorch wrapper for high-performance research, 2024

    PyTorch Lightning. Pytorch lightning: The lightweight pytorch wrapper for high-performance research, 2024. Accessed: 04-10-2024

  50. [58]

    Optuna: A hyperparameter optimization framework, 2024

    Optuna. Optuna: A hyperparameter optimization framework, 2024. Accessed: 04-10-2024

  51. [59]

    Darts: A python library for easy time series forecasting, 2024

    Darts. Darts: A python library for easy time series forecasting, 2024. Accessed: 04-10-2024

  52. [60]

    Tno security gateway, 2024

    TNO — Innovation for life. Tno security gateway, 2024

  53. [61]

    Fastapi: Modern, fast (high-performance) web framework for python, 2024

    FastAPI. Fastapi: Modern, fast (high-performance) web framework for python, 2024. Accessed: 04-10-2024

  54. [62]

    Pandas: Powerful data analysis and manipulation tool, 2024

    Pandas. Pandas: Powerful data analysis and manipulation tool, 2024. Accessed: 04-10-2024. 20 Illustrative Example inputs Supplementary Figures Figure 7: AI4EF - Training Playground - ML Pipeline 21 Figure 8: AI4EF - Excel sheet used to calculate energy savings and greenhouse g...

  55. [64]

    com/epu-ntua/ enershare-ai4ef/ blob/master/ README.md Support email for questions atzortzis@epu

    for AI4EF fron- tend and backend Link to user manual https://github. com/epu-ntua/ enershare-ai4ef/ blob/master/ README.md Support email for questions atzortzis@epu. ntua.gr 23 Table 2: MLApp Input/Output Parameters Building Retrofitting Photovoltaic Installation Inputs Target...

  56. [2023]

    Accessed: 04-10-2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.