The proliferation of
urban sensing, IoT, and big data in buildings, cities, and urban areas provides
unprecedented opportunities to better understand and optimize transportation, energy and
water networks, and how human behavior affects them (and is, in turn, affected by them).
However, historically due to poor-quality data, limitations in algorithms, and computational
bottlenecks, modeling urban-scale occupant behavior and its interactions with energy and
transportation demand has proven to be quite challenging. Therefore, progress in developing
data-driven techniques, which can work with enormous amounts of data that is increasingly
available today, is needed to unlock its full potential.
To realize this potential, BALANCES focuses on innovative data-driven methodologies for modeling
and optimizing buildings and cities.
The workshop aims to foster discussions on:
1. Big data modeling paradigms that could be applicable in building and urban science,
2. Data collection infrastructure requirements for these modeling paradigms,
3. Challenges faced by current modeling approaches, and
4. Future research directions to fully utilize building and urban big data.
BALANCES'26 will be held in conjunction with
ACM
BuildSys 2026 and the ACM e-Energy within ACM Sustainability Week.
The workshop will accept
the submissions of original work or work in progress. Submitted papers must be unpublished and
must not be currently under review for any other publication. Submissions must be full papers,
at most 4 single-spaced US Letter (8.5" x 11") pages, including figures, tables, references and
appendices. Submissions for All submissions must use the LaTeX (preferred) or Word styles found
here. All submissions must be submitted using the submission
website. Authors must make a good faith effort to anonymize their submissions by: (1)
using the "anonymous" option for the class and (2) using "anonsuppress" section where
appropriate. Papers that do not meet the size, formatting, and anonymization requirements will
not be reviewed. 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 tentative presentation formats are regular oral presentation
(15 minutes) and a spotlight presentation (2 minutes).
Register through ACM
BuildSys 2026
Please submit through
https://acmbalances26.hotcrp.com/.
And register through
https://buildsys.acm.org/2026/.
Prof. Na Li
Harvard University
9:30 Presentation 1: Beyond Prototypes: Overcoming the UBEM Data Bottleneck with Large Multimodal Models
R. Dubois, B. Howard
9:45 Presentation 2: Uncertainty-Aware Transfer Learning for Cross-Building Energy Forecasting: Toward Robust and Scalable District-Level Energy Management
S. Zaregarizi, K. Yavari
10:00 Presentation 3: An AI Agent Framework for Real-World Chiller Plant Optimization with Long-Term Deployment
J. Zhang, Z. Chen, Z. Xiao, K. Xu, T. Ma, F. Xiao
10:15 Presentation 4: Extreme Heat and Human Mobility: A Diffusion Model Approach to Climate-Adaptive Urban Dynamics
Y. Yang, B. Dong
10:30 Presentation 5: Enhancing Urban Building Energy Models through Activity Based Occupancy Generation
J. Schiller, L. Friedrich, M. Pruckner
11:00 Presentation 6: SingSys: Acoustic Vibration Sensing for Strength Monitoring and Characterization of Engineered Biopolymer Composite (EBC)
B. Miao, A. Theissler, Y. Dong, A. Lesh, H. Noh, D. Loftus, M. Lepech
11:15 Presentation 7: Bayesian-Calibrated Posterior-Ensemble Reinforcement Learning for Robust HVAC Control under Model Uncertainty
C. Shen, S. Lee, C. Lee
11:30 Presentation 8: Cross-Task Federated Backbone Aggregation with Selective State Space Models for Building Energy Analytics
B. Kumar, N. Srivastava, P. Arjunan
11:45 Presentation 9: Exploring the Potential of Physics-Informed Neural Networks for Building Energy Modeling with Component-Level Calibration
Y. Cai, A. Aryal
12:00 Presentation 10: Generalizing HVAC Control With Domain Randomized Reinforcement Learning
P. Boitel, K. Zhang
Carnegie Mellon University
USA
Colorado School of Mines
USA
National University of Singapore
Singapore
Pacific Northwest National Laboratory
USA
Indian Institute of Science
India
Oak Ridge National Laboratory
USA
University of Würzburg
Germany
Texas A&M University
USA
University of Central Florida
USA
Konkuk University
South Korea
Arizona State University
USA
Kennesaw State University
USA