Digital Twins enabled Building Automation System for comfortable, healthy and energy efficient buildings
Lead Research Organisation:
University of Cambridge
Department Name: Engineering
Abstract
As Europeans spend approximately 90% of their time indoors, the quality of confined spaces is a major concern for healthy indoor environments in Europe and it has a decisive impact on people's health and comfort and productivity. Indoor air quality, thermal and acoustic comfort and sufficient levels of lighting are the major determinants of the indoor environmental quality (IEQ) and play an
important role in ensuring the quality of life and general wellbeing of building occupants TwinBAS seizes on the timely opportunity arising from the new revised Energy Performance of Buildings Directive (EPBD, the European Green Deal, and the need for green recovery from the COVID-19 outbreak, to support the transition towards 'smarter' buildings with a long-term objective of creating
buildings that are comfortable and healthy for the occupants yet also energy efficient. In order to harness the potential benefits of recent technological advancements, the proposed work proposes a holistic assessment of tertiary sector buildings, resting on three main pillars: (i) indoor environmental conditions (ii) energy performance and (iii) the smartness of the building. Specifically, Digital
Twins will be developed to improve IEQ within building spaces based on both individual and collaborative user-preferences, by using real data processing with machine learning techniques and hybrid models that combines physics-based simulations. The Digital Twins enabled system will automatically inform control functions for optimizing building operations to maximise comfort and minimize energy use at each timestep. TwinBAS will be conducted at Cambridge University (UK), Engineering Department, where the Asset Management research group is currently conducting extended research focused on the development and exploitation of digital twins to improve asset management. A secondment of 4 months is also planned at the Integrated Environmental Solutions, Research and Development Ltd. (IES R&D) in Ireland
important role in ensuring the quality of life and general wellbeing of building occupants TwinBAS seizes on the timely opportunity arising from the new revised Energy Performance of Buildings Directive (EPBD, the European Green Deal, and the need for green recovery from the COVID-19 outbreak, to support the transition towards 'smarter' buildings with a long-term objective of creating
buildings that are comfortable and healthy for the occupants yet also energy efficient. In order to harness the potential benefits of recent technological advancements, the proposed work proposes a holistic assessment of tertiary sector buildings, resting on three main pillars: (i) indoor environmental conditions (ii) energy performance and (iii) the smartness of the building. Specifically, Digital
Twins will be developed to improve IEQ within building spaces based on both individual and collaborative user-preferences, by using real data processing with machine learning techniques and hybrid models that combines physics-based simulations. The Digital Twins enabled system will automatically inform control functions for optimizing building operations to maximise comfort and minimize energy use at each timestep. TwinBAS will be conducted at Cambridge University (UK), Engineering Department, where the Asset Management research group is currently conducting extended research focused on the development and exploitation of digital twins to improve asset management. A secondment of 4 months is also planned at the Integrated Environmental Solutions, Research and Development Ltd. (IES R&D) in Ireland
Publications
Karatzas S
(2024)
A virtual sensing approach to enhancing personalized strategies for indoor environmental quality and residential energy management
in Building and Environment
| Description | The TwinBAS project accomplished several key milestones, particularly in advancing the integration of Digital Twins into Building Automation Systems (BAS) to enhance indoor environmental quality (IEQ) and energy efficiency. Some notable achievements include: Development of hybrid digital twin technology incorporating machine learning to optimize comfort and energy performance. Successful implementation of an innovative sensing network for continuous IEQ monitoring. Introduction of algorithms to personalize comfort management rather than relying on the traditional "one-size-fits-all" approach. The project met most of its core objectives, including: Integration of environmental parameters, comfort perception, and control actions into a common normative model - Achieved through the development of a data-driven model. Enabling low-cost and continuous IEQ monitoring - Implemented via an innovative sensor network. Shifting from conventional comfort management to personalized solutions - Addressed with machine learning models. Leveraging an autonomous hybrid digital twin for optimal building control - Demonstrated in real testbeds. Validation and real-world deployment: Although tested in a controlled environment, scaling up the system for diverse buildings requires further research. Energy-comfort trade-offs: Maintaining energy efficiency while ensuring individual comfort levels remains a challenge, necessitating further refinement of predictive models. |
| Exploitation Route | The results of TwinBAS provide a strong foundation for future work in smart building automation and digital twin technology. Potential ways to take the findings forward include: Further research and development: Universities and research institutions can build upon the TwinBAS framework to refine predictive models and expand use cases. Industry adoption: Companies specializing in building automation, such as smart HVAC and sensor manufacturers, can integrate the developed methodologies into their products. Policy and regulation: Governments and regulatory bodies can use the findings to shape standards for smart, energy-efficient buildings. Commercial applications: Facility managers can deploy TwinBAS-based systems to optimize building performance and occupant well-being |
| Sectors | Construction |
| Description | The findings from TwinBAS can influence different sectors in the following ways: Public Sector: The project contributes to European policy development by informing updates to building energy efficiency standards and regulations (e.g., Smart Readiness Indicator (SRI) and European Building Performance Directive (EPBD)). Collaboration with the Ideal European cluster Private Sector: Engaged in business outreach through seminars and partnerships with industry players like REHVA (Federation of European Heating, Ventilation, and Air Conditioning Associations) and the International Society of Automation. Development of AI-driven comfort and energy management solutions offers a competitive edge to businesses operating in real estate, facility management, and HVAC industries. |
| First Year Of Impact | 2024 |
| Sector | Construction |
| Impact Types | Policy & public services |
| Description | Development of material for Digital Twins postgraduate course |
| Geographic Reach | National |
| Policy Influence Type | Influenced training of practitioners or researchers |
| Impact | New skills developed for researchers and new knowledge delivered to improve healthy building environments |
| Title | Optimizing Residential Energy Management: Integrating Air Quality and Consumer Preferences |
| Description | Mixed-Integer Linear Programming (MILP) framework is introduced ,aimed to minimize the daily energy costs associated with controllable electricity loads in residential settings while accommodating consumer's preferences regarding scheduling, thermal comfort and air quality status. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2024 |
| Provided To Others? | No |
| Impact | A key novelty of this framework is the methodical incorporation of the IAQ index as an explicit constraint within the objective function of the optimization problem. This pivotal integration catalyzes the efficient operation of the Air Purifiers units, adeptly maintaining healthy indoor air conditions without forfeiting energy efficiency, indicating a substantial reduction in daily electricity costs, achieving a 24.68% decrease compared to a baseline scenario. |
| Title | Virtual Sensors algorithms |
| Description | virtual sensing framework addressing two critical aspects of IEQ: indoor thermal comfort and indoor air quality (IAQ). |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2025 |
| Provided To Others? | Yes |
| Impact | This model is trained on a dataset including VOC, PM2.5, indoor air temperature, indoor relative humidity, outdoor air temperature and occupancy status. Specifically, occupancy status is a binary variable indicating human presence in the residence, which is a critical determinant for IAQ index forecasting.The dataset is apportioned into training (70%), validation (15%), and testing (15%) subsets, aligning with best practices to validate the model's extrapolative strength and ensure that it remains an unbiased estimator of the underlying IAQ dynamics. In evaluating the predictive performance of the LSTM network for day-ahead forecasting, the model exhibits a considerable level of accuracy. |
| Description | Smart wristbands for biosensing |
| Organisation | University of Patras |
| Country | Greece |
| Sector | Academic/University |
| PI Contribution | The wristband has been developed by University of Patras within TwinERGY H2020 project, and is being deployed in the study experiment, testing it by extending the context to include comfort apart from energy efficiency. |
| Collaborator Contribution | NA |
| Impact | Smart wristbands testing and validation |
| Start Year | 2023 |
| Description | University Visit |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Postgraduate students |
| Results and Impact | 3 postgraduate and 4 professors participated at a discussion about the research objectives and outcomes so far, and they expressed interest for future collaboration |
| Year(s) Of Engagement Activity | 2023 |