Desjardins Living Lab: Building Retrofit from Design to Implementation Towards Prosumer and Grid-Support Services
Project overview
This project is a collaboration with Groupe Desjardins to develop a digitized decarbonization roadmap using its extensive real estate holdings in Quebec, Canada and North America, focusing on showcasing smart building transformations.
As a prominent Canadian financial institution, Desjardins is pivotal in guiding Canada's buildings towards zero emissions by establishing guidelines for smart operations and retrofits. The project aims to overcome the inconsistency in decarbonization efforts by setting clear methodologies and KPIs, integrating smart grid and building operations for evolving energy needs.
We plan to improve decision-making with digital tools and IoT data for immediate and strategic enhancements and retrofit existing buildings into intelligent ones for efficient energy management. Supported by existing collaborations, the initiative seeks to lead innovative research in building transformation, involving the academic community and diverse partners.
Key project details
| Principal investigator | Manar Amayri, assistant professor, Concordia Institute for Information Systems Engineering, Concordia University |
Co-principal investigators |
Yann-Gaël Guéhéneuc, professor, Computer Science and Software Engineering, Concordia University; Nizar Bouguila, professor, Concordia Institute for Information Systems Engineering, Concordia University |
Research collaborators |
Ursula Eicker, Concordia University; Bernard Bitar, Groupe Desjardins; Mayi Kato, Groupe Desjardins |
| Non-academic partners | Groupe Desjardins |
| Research Keywords | internet-of-things, electric vehicles, distributed energy resources, building energy management systems, ai-based load monitoring |
| Budget | Cash: $240,000 In-Kind: $40,000 |
Publications:
O. Bouhamed, M. Dissem, M. Amayri, and N. Bouguila, “Transformer-based deep probabilistic network for load forecasting,” Engineering Applications of Artificial Intelligence, vol. 152, p. 110781, Jul. 2025, doi: 10.1016/j.engappai.2025.110781.
S. B. Brahim, M. Amayri, and N. Bouguila, “One-day-ahead electricity load forecasting of non-residential buildings using a modified Transformer-BiLSTM adversarial domain adaptation forecaster,” Int. J. Dynam. Control, vol. 13, no. 5, p. 176, May 2025, doi: 10.1007/s40435-025-01701-x.
B. M. Fahim, M. K. Akbar, and M. Amayri, “ResiDualNet: A novel electric vehicle charging data imputation technique to enhance load forecasting accuracy,” Build. Simul., vol. 18, no. 4, pp. 897–922, Apr. 2025, doi: 10.1007/s12273-025-1236-8.
S. Kallel, M. Amayri, and N. Bouguila, “Clustering and Interpretability of Residential Electricity Demand Profiles,” Sensors, vol. 25, no. 7, p. 2026, Mar. 2025, doi: 10.3390/s25072026.
J. Guo, W. Fan, M. Amayri, and N. Bouguila, “Deep clustering analysis via variational autoencoder with Gamma mixture latent embeddings,” Neural Networks, vol. 183, p. 106979, Mar. 2025, doi: 10.1016/j.neunet.2024.106979.
O. Sghaier, M. Amayri, and N. Bouguila, “Dirichlet and Liouville-based normality scores for deep anomaly detection using transformations: applications to images and beyond images,” Appl Intell, vol. 55, no. 1, p. 25, Jan. 2025, doi: 10.1007/s10489-024-05892-2.
M. Dissem and M. Amayri, “Unsupervised anomaly detection and imputation in noisy time series data for enhancing load forecasting,” Appl Intell, vol. 55, no. 1, p. 11, Jan. 2025, doi: 10.1007/s10489-024-05856-6.
J. Dridi, M. Amayri, and N. Bouguila, “Multi-Source Domain Adaptation Using Ambient Sensor Data,” Applied Artificial Intelligence, vol. 38, no. 1, p. 2429321, Dec. 2024, doi: 10.1080/08839514.2024.2429321.
O. Bouarada, M. Azam, M. Amayri, and N. Bouguila, “Hidden Markov models with multivariate bounded asymmetric student’s t-mixture model emissions,” Pattern Anal Applic, vol. 27, no. 4, p. 117, Dec. 2024, doi: 10.1007/s10044-024-01341-5.
N. Mahamoodally, J. Dridi, and M. Amayri, “Explainable domain adaptation for imbalanced occupancy estimation,” Journal of Building Engineering, vol. 97, p. 110613, Nov. 2024, doi: 10.1016/j.jobe.2024.110613.
O. Sghaier, M. Amayri, and N. Bouguila, “Libby-Novick Beta-Liouville Distribution for Enhanced Anomaly Detection in Proportional Data,” ACM Trans. Intell. Syst. Technol., vol. 15, no. 5, pp. 1–26, Oct. 2024, doi: 10.1145/3675405.
S. Chouchene, M. Amayri, and N. Bouguila, “Sparse coding-based transfer learning for energy disaggregation,” Energy and Buildings, vol. 320, p. 114498, Oct. 2024, doi: 10.1016/j.enbuild.2024.114498.
A. Rebei, M. Amayri, and N. Bouguila, “Affinity-Driven Transfer Learning for Load Forecasting,” Sensors, vol. 24, no. 17, p. 5802, Sep. 2024, doi: 10.3390/s24175802.
S. Tabarsaii, M. Amayri, N. Bouguila, and U. Eicker, “Non intrusive load monitoring using additive time series modeling via finite mixture models aggregation,” J Ambient Intell Human Comput, vol. 15, no. 9, pp. 3359–3378, Sep. 2024, doi: 10.1007/s12652-024-04814-x.
H. Al-Bazzaz, M. Azam, M. Amayri, and N. Bouguila, “Explainable finite mixture of mixtures of bounded asymmetric generalized Gaussian and Uniform distributions learning for energy demand management,” ACM Trans. Intell. Syst. Technol., vol. 15, no. 4, pp. 1–26, Aug. 2024, doi: 10.1145/3653980.
S. Samareh Abolhassani, A. Zandifar, N. Ghourchian, M. Amayri, N. Bouguila, and U. Eicker, “Occupant counting model development for urban building energy modeling using commercial off-the-shelf Wi-Fi sensing technology,” Building and Environment, vol. 258, p. 111548, Jun. 2024, doi: 10.1016/j.buildenv.2024.111548.
J. Guo, M. Amayri, W. Fan, and N. Bouguila, “A scaled dirichlet-based predictive model for occupancy estimation in smart buildings,” Appl Intell, vol. 54, no. 11–12, pp. 6981–6996, Jun. 2024, doi: 10.1007/s10489-024-05543-6.
M. K. Akbar, M. Amayri, N. Bouguila, B. Delinchant, and F. Wurtz, “Evaluation of regression models and Bayes-Ensemble Regressor technique for non-intrusive load monitoring,” Sustainable Energy, Grids and Networks, vol. 38, p. 101294, Jun. 2024, doi: 10.1016/j.segan.2024.101294.
M. Dissem, M. Amayri, and N. Bouguila, “Neural Architecture Search for Anomaly Detection in Time-Series Data of Smart Buildings: A Reinforcement Learning Approach for Optimal Autoencoder Design,” IEEE Internet Things J., vol. 11, no. 10, pp. 18059–18073, May 2024, doi: 10.1109/JIOT.2024.3360882.
J. Dridi, M. Amayri, and N. Bouguila, “Unsupervised clustering-based domain adaptation for estimating occupancy and recognizing activities in smart buildings,” Journal of Building Engineering, vol. 85, p. 108741, May 2024, doi: 10.1016/j.jobe.2024.108741.
A. Rebei, M. Amayri, and N. Bouguila, “FSNet: A Hybrid Model for Seasonal Forecasting,” IEEE Trans. Emerg. Top. Comput. Intell., vol. 8, no. 2, pp. 1167–1180, Apr. 2024, doi: 10.1109/TETCI.2023.3290050.
K. Maanicshah, M. Amayri, and N. Bouguila, “Novel mixture allocation models for topic learning,” Computational Intelligence, vol. 40, no. 2, p. e12641, Apr. 2024, doi: 10.1111/coin.12641.
Accepted publications in national and international conferences:
Z. Luo, W. Fan, M. Amayri, and N. Bouguila, “Dynamic Deep Clustering of High-Dimensional Directional Data via Hyperspherical Embeddings with Bayesian Nonparametric Mixtures,” in Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1, Toronto ON Canada: ACM, Jul. 2025, pp. 938–949. doi: 10.1145/3690624.3709230.
S. Hong, F. Najar, M. Amayri, and N. Bouguila, “Flexible Dirichlet Mixture Model for Multi-modal data Clustering,” in The International FLAIRS Conference Proceedings, May 2025. doi: 10.32473/flairs.38.1.138970.
F. Alkhawaja, M. Amayri, and N. Bouguila, “A hierarchical count data clustering based on Multinomial Nested Dirichlet Mixture using the Minorization-Maximization framework,” in The International FLAIRS Conference Proceedings, May 2024. doi: 10.32473/flairs.37.1.135263.
M. Amayri and N. Bouguila, “Simultaneous count data feature selection and clustering using Multinomial Nested Dirichlet Mixture,” May 2024.
Research focus
Sustainable action plan development
This project will leverage the data and experience of Desjardins to formulate a sustainable action plan for retrofitting and decision-making processes on a large scale.
Building stock management and retrofit scenario analysis
This goal aims to develop a validated methodology and set of Key Performance Indicators (KPIs) for managing and prioritizing retrofit measures for building portfolios, including decommissioning steps if necessary.
Living lab demonstration
The project will use the Complexe Desjardins (450 De Maisonneuve Blvd) in downtown Montreal as a living lab to demonstrate the feasibility of retrofitting commercial buildings according to sustainability practices and their potential to offer grid services.
End-to-end management methodology for decarbonization
This project aims to provide a comprehensive methodology for the management of real-estate that facilitates the transition to a decarbonized and sustainable building stock, including the introduction of new services from the decision-making stage to operation.
Building intelligence, metering and monitoring analysis
This project will establish a step-by-step process for transforming existing buildings into intelligent structures with Distributed Energy Resources (DERs), metering and monitoring capabilities.
Integration of optimization and communication for self-consumption and grid services
This step aims to develop intelligent algorithms (optimization and AI-based) to manage building systems for self-consumption and support of grid services with the necessary metering and communication technology.
Non-academic partners
Thank you to our non-academic partners for your support and trust.
Volt-Age is funded by a $123-million grant from the Canada First Research Excellence Fund.
