FMSust 2025

The first ACM International workshop on Foundation Models for Energy-Efficient Buildings, Cities, Transportation, and Sustainability (FMSust)
(A BuildSys 2025 Workshop)
November 19, 2025
Golden, Colorado, USA

About FMSust 2025

The rapid advancement of foundation models offers a transformative opportunity to address some of the most pressing challenges in sustainable development. Unlike traditional machine learning models that are typically designed for single-task solutions, foundation models – especially Large Language Models (LLMs), Vision-Language Models, Multimodal Foundation Models, and Time Series Foundation Models (TSFMs) – are large-scale models trained on vast and diverse datasets. They are designed to handle multiple downstream tasks and offer high generalizability and adaptability.

In recent years, there has been growing interest in developing domain-specific foundation models to address the challenges of building and deploying these models in specialized contexts. Within the SIGEnergy community, there is growing momentum to develop, analyze, and explore the capabilities and limitations of such models, and to assess their adaptability across a range of tasks. The 1st International Workshop on Foundation Models for Energy-Efficient Buildings, Cities, Transportation, and Sustainability provides a timely platform for researchers and industry practitioners to exchange ideas and share their latest findings, with the goal of advancing our collective understanding and responsible use of foundation models in the energy and sustainability domains.

Call for Papers

FMSust 2025 Call for Papers (PDF)
Topics of interest for the workshop include (but are not limited to) the following:

Foundation Models for Energy and Sustainability

  • Domain-specific foundation models for buildings, cities, grids, and transportation
  • Large Language Models (LLMs) and their applications in energy and sustainability
  • Vision-language models for multimodal understanding and control
  • Multimodal foundation models combining text, vision, time series, and structured data
  • Time Series Foundation Models (TSFMs) for forecasting, control, and anomaly detection
  • Transfer learning, fine-tuning, and adaptation techniques
  • Efficient and scalable model architectures for low-resource or edge environments
  • Modeling and optimization for decarbonization pathways

Datasets, Benchmarking, and Evaluation

  • Creation and curation of domain-specific datasets for training foundation models
  • Open benchmarks and reproducibility frameworks
  • Standardized metrics for evaluating generalization, fairness, and robustness
  • Tools and platforms for dataset sharing, annotation, and synthetic data generation

Deployment, Validation, and Impact Assessment

  • Real-world deployment of foundation models
  • Field studies, pilot projects, and validation of foundation models
  • Socio-technical challenges in deploying foundation models at scale
  • Monitoring, adaptation, and continual learning from deployed systems
  • Assessment of environmental and social impacts of model use

Cross-Cutting Themes

  • Human-in-the-loop systems and interpretable foundation models
  • Ethical, legal, and responsible AI considerations
  • Privacy and security issues
  • Federated foundation models
  • Democratizing access to foundation models

Submission Guidelines

The workshop solicits submissions for technical papers, or works in progress, reporting on novel research to be presented at the workshop (in person). Submitted papers must be unpublished and must not be currently under review for any other publication. Paper submissions must be at most 4 single-spaced US Letter (8.5"x11") pages, including figures, tables, and appendices (excluding references). All submissions must use the LaTeX (preferred) or Word styles found here. Please note that ACM uses 9-pt fonts in all conference proceedings, and the style (both LaTeX and Word) implicitly define the font size to be 9-pt. The workshop will follow a single-blind review process. Authors must disclose their identities in the paper, while reviewers’ identities will remain anonymous.


Submission link

All submissions must be in Adobe Portable Document Format (PDF) format through the HotCRP: https://buildsys25-fmsust.hotcrp.com/

Important Dates

  • Paper submission: September 22, 2025 (AOE)
  • Notifications: September 30, 2025 (AOE)
  • Camera-ready: October 3, 2025 (AOE)
  • Workshop: November 19, 2025

Program

*All times are local (Golden, Colorado, USA)(UTC−6)

8:15 – 8:30
Opening Remarks
8:30 – 9:15
Keynote Talk (jointly with DataFM)

Opportunities and Challenges with Foundation Models for Smart Environments and Energy Sustainability
Pandarasamy Arjunan (Indian Institute of Science)

9:15 – 10:15
Technical Session (FMSust and CPSIS papers)
  • Tackling Non-Stationarity in HVAC Control with TimeGPT-Enhanced Deep Reinforcement Learning
    Jiatong Li (Xidian University); Bokai Ji, Guangxia Li (Xidian University); Peilin Zhao (Shanghai Jiao Tong University); Liu Liu (Tencent); Zhigang Ren (Xian Huayun Wisdom Information Technologies Co., Ltd., China)
  • LLM-Driven Agent-Based Evaluation Framework for Occupant-Centric HVAC Control Considering Individual Differences and Movement Behavior
    Shinnosuke Sasaki (The University of Osaka); Nattaon Techasarntikul (The University of Osaka); Yuichi Ohsita (The University of Osaka); Hideyuki Shimonishi (The University of Osaka)
  • SifterNet: Model-Agnostic Defense against Backdoor Attack in Vision Large Model
    Shaoye Luo, Xinxin Fan (University of Chinese Academy of Sciences, Institute of Computing Technology, Chinese Academy of Sciences); Quanliang Jing (Institute of Computing Technology, Chinese Academy of Sciences); Men Niu (University of Chinese Academy of Sciences, Institute of Computing Technology, Chinese Academy of Sciences); Chi Lin (Dalian University of Technology); Yunfeng Lu (Beihang University)
  • Defining Social Capital and Social Infrastructure for Built Environment Design
    Kieran Elrod (Carnegie Mellon University)
10:15 – 10:30
Coffee Break
10:30 – 11:30
Invited Talks
  • Towards Zero-shot Question Answering in CPS-IoT: Large Language Models and Knowledge Graphs
    Ozan Baris Mulayim (Carnegie Mellon University; Lawrence Berkeley National Laboratory), Gabe Fierro (Colorado School of Mines), Mario Bergés (Carnegie Mellon University), Marco Pritoni (Lawrence Berkeley National Laboratory)
  • Exploring Capabilities of Time Series Foundation Models in Building Analytics
    Xiachong Lin (University of New South Wales, Australia), Arian Prabowo (University of New South Wales, Australia), Imran Razzak (University of New South Wales, Australia), Hao Xue (University of New South Wales, Australia), Matthew Amos (CSIRO Energy Centre, Australia), Sam Behrens (CSIRO Energy Centre, Australia), Flora D. Salim (University of New South Wales, Australia)
  • From RNNs to Foundation Models: An Empirical Study on Commercial Building Energy Consumption
    Shourya Bose (UC Santa Cruz), Yijiang Li (Argonne National Laboratory), Amy Van Sant (National Renewable Energy Laboratory), Yu Zhang (UC Santa Cruz), Kibaek Kim (Argonne National Laboratory)
  • Global Carbon Intensity Forecasting Through Time Series Foundation Models
    Diptyaroop Maji (University of Massachusetts Amherst), Kang Yang (University of California Los Angeles), Prashant Shenoy (University of Massachusetts Amherst), Ramesh K. Sitaraman (University of Massachusetts Amherst), Mani Srivastava (University of California Los Angeles)
11:30 – 11:45
Closing Remarks

Organization

General Co-chairs

Stephen, Lee
Stephen, Lee
Department of Computer Science
University of Pittsburgh
Pandarasamy Arjunan
Pandarasamy Arjunan
Indian Institute of Science, Bangalore
India
Prashant Shenoy
Prashant Shenoy
College of Information and Computer Sciences
University of Massachusetts

Registration and Venue

For registration and venue details, visa information, etcetera please visit the BuildSys webpage