Buildings consume one-third of the world's energy and are some of the major energy consumers on the planet. Occupants use this energy for enhancing indoor environmental quality (IEQ) which is afected by many factors including temperature, humidity, airflow, air quality, etc.; however, it is dificult to find a suitable general solution to improve IEQ while decreasing energy usage as each building is under diferent environmental conditions, and every occupant has diferent clothing insulation and a diferent metabolic rate. In this work, we propose an ontology that, based on real-time data from Internet of Things (IoT), can be used to suggest a viable solution to enhance IEQ and decrease energy usage by combining several sets of knowledge: indoor environmental conditions, outdoor environmental conditions, and occupant profiles. We demonstrated that the ontology with an application can recommend suitable actions to improve IEQ. In future work, this ontology could serve as the foundation on top of which to develop an occupant-centric building automation system for enhancing IEQ and energy eficiency, based on 3D geometric models and thermodynamic simulation modules.
According to reports from the U.S. Energy Information Administration (US EIA), commercial
and residential buildings were responsible for 72% of electric energy in 2013 [
In a given room, thermal comfort is afected by many factors: air temperature, mean radiant
temperature, relative humidity, airflow, air quality, clothing, human activity, and occupant
information such as age, sex, height, and weight [
In this work, we approach this problem from a Semantic standpoint by proposing an ontology that can be used to find a viable solution to improve IEQ for occupants while minimizing energy use in a room. The ontology combines several sets of knowledge: 1) indoor environmental conditions including air temperature, relative humidity, air speed, sound pressure, and luminosity, 2) outdoor environmental conditions, such as air quality and daylight intensity, and 3) occupant information including sex, height, weight, age, clothing insulation, and activity level.
We assessed the ontology through a set of simulated scenarios where user profiles can be used to quantify the IEQ requirements and evaluate the IEQ management system with competency questions. Users can inform quantitative factors for thermal comfort, acoustic comfort, visual comfort, and air quality that are currently causing them discomfort and to what degree and the system will suggest a method for bringing those factors into an acceptable range. The user can manually enter their desired temperature and humidity ranges, or the system can infer them through the Predicted Mean Vote (PMV) model based on IoT sensor data as well as other information that the user provides, including occupant profile descriptions. We found that the ontology was able to be used to infer how to change indoor environmental parameters to meet the comfort requirements of multiple occupants and suggest viable actions to reach the desirable indoor environmental condition.
The goal of this ontology is to provide suggestions to improve IEQ in a room based on indoor
and outdoor environments and occupant profiles from IoT sensors or users. To evaluate thermal
comfort, this ontology utilizes the PMV model standardized by International Organization for
Standardization (ISO) and Air Quality Index (AQI) established by the United States Energy
Information Administration (US EPA). Calculating the PMV index requires air temperature,
air speed, relative humidity, clothing insulation, and metabolic rate [
To constrain the coverage of the ontology, we focused on several usage scenarios involving indoor and outdoor environmental conditions and occupants’ demographic information. Based on the requirements and competency questions that we extracted from these usage scenarios, we further developed the key concepts and relations necessary in our ontology.
To suggest an action for increasing the occupants’ overall comfort, our ontology-enabled system supports this reasoning by connecting a room and its components, occupant profiles, and indoor and outdoor environmental parameters. The primary parameters that this system acts on are environmental measurements taken by available IoT sensors.
The figure below shows the relationships between the most important high-level resources in our
ontology. Central to our project is a Room, which has Room Component—objects in the room that
have some efect on the room’s environment—and one or more Occupants, which have various
characteristics from which we may calculate a comfort range. Room Component—objects are
either power-consuming or non-power-consuming, with priority given to actions that use
NonPower Consuming Component—objects during action recommendations. Each Room Component
has multiple possible Component States and Component Actions; each action produces a new
Component State, as well as a diferent Environment consisting of IEQ parameters measured by
saref:Sensor from SAREF ontology[
Based on existing ontologies including SAREF[
For the scope of this project, we designed six competency questions to evaluate the eficacy of the ontology. The questions mainly focus on asking for a strategy to enhance occupants’ comfort by changing given indoor and outdoor environmental parameters. These are complex problems because the occupants’ comfort depends not only on indoor environmental parameters but also on occupant profiles. Furthermore, solutions can be diferent depending on available room components and outdoor environmental parameters as well as the indoor environment. In this paper, we show only two representative competency questions due to the limited number of pages. The full set of the questions can be found on our website.
We performed assessments by constructing SPARQL queries and verifying the answer to
each question. Note that this evaluation is carried out only for assessing the ability to answer
the questions through manual inputs, and does not cover the capability of an IEQ management
system using this ontology. Moreover, because our application uses OWL (DL) reasoning but
cannot perform arithmetic or numeric comparison, we assume that users either directly input
their comfort range or permit that it be calculated using a PMV equation [
Additionally, we assume that users manually input their activity level and clothing insulation
based on standardized data tables in ANSI/ASHRAE Standard [
Question: How should IEQ parameters, such as temperature, humidity, airflow, etc., be changed to make the multiple occupants feel comfortable in a living room during summer? The occupants’ profiles are a 26-year-old daughter typing something on his laptop, a 59-year-old mother dancing, and a 32-year-old son cleaning the house. The outdoor weather is 89°F, relative humidity is 70%, and the outdoor air quality index is 34, “Good”. Indoor temperature is 85°F and relative humidity is 67%. A fan and a dehumidifier are available.
This query looks for two diferent actions: one to change the air temperature and the other to change the relative humidity. Each action must be available for a particular room component that’s, in turn, part of the room individual that’s associated with the relevant competency question. The actions are selected by ensuring that they produce respective resultant environments with the same environment attribute delta signs as the target environment.
Question: In a small gym, three people are working out. 22-year-old male Jason is walking on a treadmill and lifting 45 kg bars with shorts & a short-sleeve shirt, 44-year-old male Bob is seated and conducting heavy limb movements with typical summer indoor clothing, and 52-year-old female Sarah is walking on a treadmill at 3 mph with a short-sleeve shirt. How should IEQ parameters, such as temperature, humidity, airflow, etc., be changed to make the multiple occupants feel comfortable in a gym? The indoor air speed is 0.3m/s, the outdoor air speed is 2m/s, and the outdoor air quality index is 38, ‘Good’. An air conditioner is available, and all windows are closed.
This query looks for a single action to change the air speed. The action must be available for a particular room component that is, in turn, part of the room individual that’s associated with the relevant competency question. An action is selected by ensuring that it produces a resultant environment with the same air speed environment attribute delta sign as the target environment. The query also requires that the resultant environment have a “good” air quality level, which is inferred by the reasoner from the fact that opening a window must produce a resultant environment with the same air quality level as the relevant outdoor environment.
Through comparative analysis for 17 articles, we classified the related works into three categories: energy management, post-occupancy evaluation (POE), and indoor environmental quality (IEQ).
The ontologies in the first category were designed to identify ineficient energy consumption
patterns and provide advice to improve eficiency; however, they don’t concern occupants’
comfort [
To overcome the limitations in the three categories of previous research works, this paper focuses on developing an ontology that provides advice to improve indoor environmental quality and reduce energy consumption based on occupants’ profiles and available room components.
We use semantics to infer how to change indoor environmental parameters to meet the comfort requirements of multiple occupants. For instance, our ontology can be used to infer that air speed should be increased, decreased, or unchanged based on the diferent comfort ranges of three occupants. Additionally, semantics can be utilized to infer whether particular actions are “acceptable” given a set of general rules and heuristics. For example, the ontology is designed such that a reasoner can infer that opening a window produces a resultant indoor environment with the same air quality level as the relevant outdoor environment. A query might then restrict the set of actions that it returns to just those that produce a “good” or “moderate” indoor air quality level. Resultant indoor environments are predicted, not detected in the real world, so a query on a regular database without semantics would not be able to filter out actions that cause unacceptable indoor air quality levels because the necessary information would not be present in the database.
The main benefits of this semantic approach over other techniques, such as machine learning or elaborative scripts, are extensibility and robustness. This approach makes it easier for existing semantic systems to integrate with our ontology, reducing the burden of adoption in smart buildings. Additionally, our approach gracefully degrades with a lack of complete inputs, which is crucial when IoT sensors go ofline or occupants fail to provide all requested information.
Firstly, the most significant limitation of our model is that an ideal indoor environment must be declared in terms of a positive or negative delta from the current indoor environment for air temperature, air speed, relative humidity, luminosity, and sound pressure, and reasoning on precise numeric values is unsupported. One notable consequence of this is that multiple actions that afect the same IEQ metric can’t be “summed” to produce a single delta of greater or lesser magnitude. Secondly, this ontology cannot consider the interrelation between parameters of thermal comfort. For example, an occupant’s thermal comfort ranges of air temperature and relative humidity depend on air speed, clothing insulation, and metabolic rate; however, our current ontology cannot fully capture this relationship. Thirdly, our model assumes that indoor environmental parameters are uniform for all locations in a room, so it could potentially suggest an improper solution if the size of the room is large and the distribution of the air temperature is uneven.
These limitations are the results of scoping decisions made at the beginning of the ontology development process for feasibility. We don’t currently foresee any specific technical hurdles that would preclude the expansion of the ontology to overcome these limitations in the future.
The main contribution of this work is to develop an ontology suggesting viable solutions for enhancing IEQ considering indoor and outdoor environmental conditions and occupant profiles. If non-power-consuming components are available, then a query on the ontology could place a higher priority on them to reduce building energy use. Additionally, we described competency questions with simulated environmental settings to specify the scope of this work and the essential functionality. We demonstrated the ontology’s functionality by answering the competency questions using SPARQL queries. Formal reasoning with semantic technologies enables the filtering out of undesirable action suggestions. Furthermore, unlike the related works, the proposed ontology includes multiple occupant profile concepts to find the optimal IEQ parameters for their diferent comfort requirements. Finally, existing semantic systems can be easily integrated with our ontology due to the extensibility of the ontological approach.
In the future, our current reasoning system using signs will be expanded to consider more granular changes in parameters by either properly considering precise numeric values computed outside of the ontology rather than positive or negative delta signs. Similarly, window-related logic should be put in place to reason about the efects of other outdoor parameters, such as temperature, on the related indoor space. Furthermore, to be practical in a large room where the available parameters will difer in diferent points, our ontology must incorporate geometric and thermodynamic reasoning for supporting multiple “sub-environments” and reasoning about what areas of an environment room components can afect. Additionally, such a system should be able to make suggestions considering spatial information and thermodynamic simulation, such as moving a fan to be closer to a certain person with lower temperature preferences.
This ontology could serve as the foundation on top of which to develop an occupant-centric building automation system for improving IEQ and reducing building energy usage, based on 3D geometric models and thermodynamic simulation modules. ontology-based framework for building energy management with iot, Electronics 8 (2019) 485. doi:10.3390/electronics8050485. [18] H. Wicaksono, P. Dobreva, P. Häfner, S. Rogalski, Methodology to develop ontological building information model for energy management system in building operational phase, in: Knowledge Discovery, Knowledge Engineering and Knowledge Management: 5th International Joint Conference, IC3K 2013, Vilamoura, Portugal, September 19-22, 2013.
Revised Selected Papers 5, Springer, 2015, pp. 168–181. doi:/978-3-662-46549-3_11. [19] H. Pruvost, A. Wilde, O. Enge-Rosenblatt, Ontology-based expert system for automated monitoring of building energy systems, Journal of Computing in Civil Engineering 37 (2023) 04022054. doi:10.1061/(ASCE)CP.1943-5487.0001065. [20] N. M. Tomašević, M. Č. Batić, L. M. Blanes, M. M. Keane, S. Vraneš, Ontology-based facility data model for energy management, Advanced Engineering Informatics 29 (2015) 971–984. doi:10.1016/j.aei.2015.09.003. [21] H. Li, T. Hong, A semantic ontology for representing and quantifying energy flexibility of buildings, Advances in Applied Energy 8 (2022) 100113. doi:10.1016/j.adapen.2022. 100113. [22] T. Hong, S. D’Oca, W. J. Turner, S. C. Taylor-Lange, An ontology to represent energy-related occupant behavior in buildings. part i: Introduction to the dnas framework, Building and Environment 92 (2015) 764–777. doi:10.1016/j.buildenv.2015.02.019. [23] T. Hong, S. D’Oca, S. C. Taylor-Lange, W. J. Turner, Y. Chen, S. P. Corgnati, An ontology to represent energy-related occupant behavior in buildings. part ii: Implementation of the dnas framework using an xml schema, Building and Environment 94 (2015) 196–205. doi:10.1016/j.buildenv.2015.08.006. [24] Y. Zhao, Q. Yang, A. Fox, T. Zhang, Ontology-based knowledge modeling of post-occupancy evaluation for green building, in: IOP Conference Series: Earth and Environmental Science, volume 495, IOP Publishing, 2020, p. 012076. doi:10.1088/1755-1315/495/1/012076. [25] Y. Zhao, Q. Yang, Development of an ontology-based semantic building post-occupancy evaluation framework, International Journal of Metrology and Quality Engineering 12 (2021) 19. doi:10.1051/ijmqe/2021019. [26] M. Nolich, D. Spoladore, S. Carciotti, R. Buqi, M. Sacco, Cabin as a home: a novel comfort optimization framework for iot equipped smart environments and applications on cruise ships, Sensors 19 (2019) 1060. doi:10.3390/s19051060. [27] J. A. Adeleke, D. Moodley, An ontology for proactive indoor environmental quality monitoring and control, in: Proceedings of the 2015 annual research conference on south African institute of computer scientists and information technologists, 2015, pp. 1–10. doi:10.1145/2815782.2815816. [28] D. Spoladore, S. Arlati, S. Carciotti, M. Nolich, M. Sacco, Roomfort: An ontology-based comfort management application for hotels, Electronics 7 (2018) 345. doi:10.3390/ electronics7120345. [29] W. Chen, K. Chena, V. Gan, J. Cheng, A methodology for indoor human comfort analysis based on bim and ontology, in: ISARC. Proceedings of the International Symposium on Automation and Robotics in Construction, volume 36, IAARC Publications, 2019, pp. 1189–1196. doi:10.22260/ISARC2019/0159. [30] D. Spoladore, S. Arlati, M. Sacco, Semantic and virtual reality-enhanced configuration of domestic environments: the smart home simulator, Mobile Information Systems 2017 (2017). doi:10.1155/2017/3185481. [31] T. A. Nguyen, A. Raspitzu, M. Aiello, Ontology-based ofice activity recognition with applications for energy savings, Journal of Ambient Intelligence and Humanized Computing 5 (2014) 667–681. doi:10.1007/s12652-013-0206-7. [32] I. Esnaola-Gonzalez, J. Bermúdez, I. Fernandez, A. Arnaiz, Eepsa as a core ontology for energy eficiency and thermal comfort in buildings, Applied Ontology 16 (2021) 193–228. doi:10.3233/AO-210245.