Ontology Engineering 2022

Final Version

Use-Case Driven Knowledge Encoding

III. Usage Scenarios

1. An office worker in San Diego, California, usually works in a one-person office. During summer, the office is too hot, 86°F, due to strong sunlight, and she wants to open the window; however, outdoor humidity is 83% and the air quality index is 273, ‘Bad’. Additionally, she can’t turn on the air conditioner because it’s too old and emits dust. As a result, her IEQ is considered low. The application may recommend pulling down blinds to block the sunlight and turn on the fan. The system shows a simulation result of how this solution can consume lower electricity compared to turning on the air conditioner. In this scenario, the office worker’s IEQ comfort range is computed outside of the ontology based on contextual information (such as clothing descriptions) that she supplies.
2. Three office workers in Chicago, Illinois, work in a school. During winter, it is difficult for them to find a suitable thermostat setting. The weather is extremely cold, 18°F. One of the workers, Michael, is 22 years old, male, and he feels warm. However, Jane, who is 53 years old, wants to increase the thermostat setting of a heater — even if she wears a thick sweater. The other worker, Tom, who is 42 years old, feels cold like Jane but dislikes the humid air from the air heater. The application may recommend keeping the thermostat setting, 75°F, and also turning on the electric heater in the room. The system shows how the three people have different thermal comfort ranges and what the optimal temperature & humidity are for each of them.
3. A person is cleaning their home living room. The activity is physically intensive, so they feel too hot by about 8°F. The current outdoor temperature is 90°F, and the current indoor temperature is about 83°F. The humidity is low, measuring 25%. An air-conditioning unit is available, but there are no other climate-control options. Setting the A/C to 75°F uses a lot of energy, but the knowledge base and ontology aren’t aware of any other options, so the system falls back on suggesting this solution.

IV. Basic Flow of Events

V. Alternate Flow of Events

VI. Use Case and Activity Diagram(s)

Figure 1: Use Case Diagram

Figure 2: System Architecture

This diagram shows how the main actors in our system interact, particularly the occupant carrying out possible interactions with the system through the interface. The interface is able to display or update information from a room’s sensor network, the recommender system itself, and the system’s database, while the underlying ontology utilizes the database and supports the recommender system.

Figure 3: Activity Flow Diagram (Normal Flow)

The above activity diagram elaborates on the interaction of system components in a normal activity flow. The user, an occupant in the room that the system is currently acting on, initiates the flow by launching the application, which displays the current status of the environment via available sensors. Additionally, the user profiles of the current occupants are requested from the database, and appropriate calculations to determine a user’s optimal environment are carried out if needed. Before running the recommender, the user may augment the calculated ideal environment with their current environmental preferences.

Initialization of the recommender system requires the support of the ontology, which creates a knowledge graph based on the current status of the database. Using inference and queries, the ontology relays a possible solution, based on an action that can be imposed on some room component, to the recommender system, which displays said solution to the user. From here, the user may take this advice and choose to run the recommender again at any point, repeating the latter steps of the system’s normal flow.     

VII. Competency Questions

1. Question: Which solution does improve indoor environmental quality and makes an occupant feel comfortable? The outdoor air temperature is 76°F, daylight is 110,000 lux through the window, and the outdoor air quality index is 36, ‘Good’. The indoor air temperature is 82°F. The occupant says that it’s 9°F too hot for their comfort. The available, configurable equipment includes currently open blinds that block the window, a ceiling fan that’s currently switched off, and a window-mounted air conditioning unit.

Sample Answer from Ontology: Pull down blinds to block the sunlight.

Terms used from Ontology: room, Indoor Environmental Quality, indoor, outdoor, air temperature, relative humidity, air speed, air quality, daylight, air quality index, air-conditioner,  window blinds, energy, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, weather data, real-time sensor values such as air temperature, (2) Calculating an occupant’s thermal comfort range based on users’ input data and the loaded data, (3) Reasoning to check viable solutions to change IEQ parameters, such as pulling up/down blinds, turning on/off the air conditioner, etc., (4) Suggesting a solution using non-power-consuming components if they are available to reduce energy consumption.

Usage scenario covered: An occupant in his/her room, who feels discomfort and wants to find an optimal solution to enhance IEQ using minimal energy

Description/Description + Ontology Usage: Based on users’ input data, the system would load BIM/weather/sensor data. The ontology would be leveraged to find out how to improve IEQ by changing indoor environmental parameters. For example, if the current air temperature is too hot for the user, it could be improved by decreasing air temperature. Then, reasoning could be applied to identify which room components can be used to change the parameters. For example, opening the window can be one of the viable solutions because it can decrease air temperature if the outdoor air temperature is lower than the indoor temperature. Based on the availability of the room components, the system could suggest non-power-consuming components to reduce electric energy.

Reasoning: Of the three configurable factors (i.e., the blinds, the fan, and the A/C unit), the blinds are non-power-consuming components, while the fan and A/C are power-consuming components. Because daylight intensity is higher than 100,000 lux, lowering the blinds in a ventilated room (which can be determined with the BIM data) would lower the indoor temperature without the energy usage that comes with turning on the A/C or fan.

2. Question: How to meet all requirements of the multiple occupants’ comfort ranges in an office room? There are three occupants who prefer temperatures in the range of 73°F to 77°F, 74°F to 78°F, and 75° to 78°F, respectively. All other factors are already ideal. The outdoor temperature is 18°F. The current HVAC thermostat setting is 75°F, which is the current indoor temperature. An electric space heater is available but is currently switched off.

Sample Answer from Ontology: Keep the thermostat setting at 75°F.

Terms used from Ontology: room, Indoor Environmental Quality, indoor, outdoor, air temperature, air-conditioner, thermostat, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, weather data, real-time sensor values such as air temperature, (2) Calculating an occupant’s thermal comfort range based on users’ input data, the loaded data, (3) Reasoning to check viable solutions to change IEQ parameters, such as changing a thermostat setting of an air conditioner, fan, etc., (4) Suggesting a solution using non-power-consuming components if they are available to reduce energy consumption.

Usage scenario covered: Three occupants in an office room, who are already comfortable

Description/Description + Ontology Usage: Based on users’ input data, the system would load BIM/weather/sensor data. The ontology would be leveraged to find out how to improve IEQ by changing indoor environmental parameters. In this case, the current air temperature meets the three occupants’ temperature preferences, and IEQ parameters don’t need to be changed. Then, the system suggests to keep the thermostat setting, and the reasoning process is terminated.

Reasoning: Current HVAC thermostat setting, 75°F, is within the comfort range of all three occupants. Therefore, the thermostat setting doesn’t have to be changed.

3. Question: How IEQ parameters, such as temperature, humidity, air speed, etc., should be changed to make the multiple occupants feel comfortable in an office room? They are working on their seats quietly. The occupants’ profiles are 26-year-old female Jane (height: 5’ 2”, weight: 121 lbs, sweat pants, long-sleeve sweatshirt: 0.74 clo, the blue area in Figure 4), 59-year-old female Megan (height: 5’ 8”, weight: 136 lbs, wearing Trousers, long sleeve-shirt: 0.61 clo, the grey area in Figure 4), and 46-year-old male John (height: 6’ 1”, weight: 189 lbs, wearing sweatpants, Jacket, Trousers, long-sleeve shirt: 0.96 clo, the purple area in Figure 4). Indoor temperature is 70°F, relative humidity is 34%, and air speed is 0.1m/s. An electric heater is available.

Sample Answer from Ontology: Recommend Megan to wear a jacket 

Terms used from Ontology: room, Indoor Environmental Quality, indoor, outdoor, air temperature, relative humidity, air speed, air quality, occupant profile, age, sex, height, weight, clothing insulation, metabolic rate, electric heater, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, weather data, real-time sensor values including air temperature, relative humidity, airflow, air quality, etc., (2) Calculating an occupant’s thermal comfort range based on users’ input data, the loaded data, and the results of PMV model (see Appendix A, Appendix B, Appendix C, and Appendix D), (3) Reasoning to check viable solutions how to change IEQ parameters, such as pulling up/down blinds, opening/closing door, window, turning on/off the air conditioner, fan, electric heater, etc., (4) Suggesting a solution using non-power-consuming components if they are available to reduce energy consumption.

Usage scenario covered: Three occupants in an office room, who are doing the same activity, feel discomfort and want to find an optimal solution to enhance IEQ without using power-consuming components to reduce energy consumption

Description/Description + Ontology Usage: Based on users’ input data, the system would load BIM/weather/sensor data, and the PMV equation, the lower and upper bound of the comfort range can be calculated for the three occupants. The ontology would be leveraged to find out how to improve IEQ by changing indoor environmental parameters. For example, if a temperature value is out of occupants’ comfort range, it could be increased or decreased to improve thermal comfort. Then, reasoning could be applied to identify which room components can be viable to change the parameters. For example, an electric heater can be one of the viable solutions because it only increases air temperature. Based on the availability of the room components, the system could suggest non-power-consuming components to reduce electric energy.

Reasoning: The current indoor air temperature and relative humidity (the red dot in Figure 4) are out of only Megan’s comfort ranges (the grey areas in Figure 4). We can turn on an electric heater to increase the indoor temperature; however, it is a power-consuming component, and Megan could feel comfortable if she increases her clothing level by wearing a jacket and changes her comfort range, as shown in Figure 4.

Figure 4: Psychrometric Chart before Megan wears a jacket (left, clothing level: 0.61 clo) and after she wears a jacket (right, clothing level: 0.96 clo)

4. 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’ profile is a 26-year-old son typing something on his laptop (metabolic rate: 1.1, Long-sleeve coveralls, t-shirt: 0.72 clo, the blue area in Figure 5), a 59-year-old mother dancing (metabolic rate: 3.4, Long-sleeve coveralls, t-shirt: 0.72 clo, the grey area in Figure 5), and a 32-year-old daughter cleaning the house (metabolic rate: 2.7, Long-sleeve coveralls, t-shirt: 0.72 clo, the purple area in Figure 5). The outdoor weather is 89°F, relative humidity is 70%, and outdoor air quality index is 34, ‘Good’. Indoor temperature is 85°F and relative humidity is 67%. A fan and a dehumidifier are available.

Sample Answer from Ontology: Turned on the fan and dehumidifier

Terms used from Ontology: room, Indoor Environmental Quality, indoor, outdoor, air temperature, relative humidity, occupant profile, age, sex, height, weight, clothing insulation, metabolic rate, fan, dehumidifier, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, weather data, real-time sensor values including air temperature, relative humidity, air quality, etc., (2) Calculating an occupant’s thermal comfort range based on users’ input data, the loaded data, and the results of PMV model (see Appendix A, Appendix B, Appendix C, and Appendix D), (3) Reasoning to check viable solutions how to change IEQ parameters, such as turning on/off the fan, dehumidifier, etc., (4) Suggesting a solution using non-power-consuming components if they are available to reduce energy consumption.

Usage scenario covered: Three occupants in a living room, who are doing different activities. Only one person feels comfortable, and they want to find an optimal solution to enhance IEQ using minimal energy

Description/Description + Ontology Usage: Based on users’ input data, the system would load BIM/weather/sensor data, and the PMV equation, the lower and upper bound of the comfort range can be calculated for the three occupants. The ontology would be leveraged to find out how to improve IEQ by changing indoor environmental parameters. For example, if a temperature value is out of occupants’ comfort range, it could be increased or decreased to improve thermal comfort. Then, reasoning could be applied to identify which room components can be viable to change the parameters. For example, an electric heater can be one of the viable solutions because it only increases air temperature. Based on the availability of the room components, the system could suggest non-power-consuming components to reduce electric energy.

Reasoning: The three people have gaps in comfort ranges of temperature and humidity due to the different activity levels, and the current air temperature and relative humidity (the red dot in Figure 5) are only in the son’s comfort range (the blue area in Figure 5). If the indoor air-speed increases and the relative humidity decreases, IEQ will meet the three people’s comfort requirements, as shown in Figure 5. In this case, turning on a fan and dehumidifier can be a solution to increase air flow and decrease relative humidity. For example, if air-speed becomes 1.5m/s, relative humidity becomes 22%, all the occupants would feel comfortable.

Figure 5: Psychrometric Chart before turning on equipment (left, air-speed: 0.8m/s, relative humidity: 67%) and after turning on (right, air-speed: 1.5m/s, relative humidity: 22%)

5. Question: In a small gym, three people are working out. 22-year-old male Jason walking on a treadmill lifting 45kg bars (metabolic rate: 4.0, wearing shorts & short-sleeve shirt: 0.36 clo, the blue area in Figure 6), 44-year-old male Bob seated with heavy limb movement (metabolic rate: 2.2, wearing typical summer indoor clothing: 0.5 clo, the grey area in Figure 6), and 52-year-old female Sarah walking on a treadmill with 3 mph (metabolic rate: 3.8, wearing a short-sleeve shirt: 0.57 clo, the purple area in Figure 6). How should IEQ parameters, such as temperature, humidity, airflow, etc., be changed to make the multiple occupants feel comfortable in a gym? Indoor air-speed is 0.3m/s, outdoor air-speed is 2m/s, and outdoor air quality index is 38, ‘Good’. An air conditioner is available, and windows are closed.

Sample Answer from Ontology: Open windows

Terms used from Ontology: room, Indoor Environmental Quality, indoor, outdoor, air temperature, relative humidity, air speed, air quality, occupant profile, age, sex, height, weight, clothing insulation, metabolic rate, air-conditioner,  fan, power consumption, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, weather data, real-time sensor values including air temperature, relative humidity, airflow, air quality, etc., (2) Calculating an occupant’s thermal comfort range based on users’ input data, the loaded data, and the results of PMV model (see Appendix A, Appendix B, Appendix C, and Appendix D), (3) Reasoning to check viable solutions how to change IEQ parameters, such as opening/closing window, etc., (4) Suggesting a solution using non-power-consuming components if they are available to reduce energy consumption.

Usage scenario covered: Three occupants in a gym, who are doing different activities, feel discomfort and want to find an optimal solution to enhance IEQ using minimal energy

Description/Description + Ontology Usage: Based on users’ input data, the system loads BIM/weather/sensor data, and the lower and upper bound of the comfort range can be calculated for the three occupants. The ontology would be leveraged to find out how to improve IEQ by changing indoor environmental parameters. For example, if a temperature value is out of occupants’ comfort range, it could be increased or decreased to improve thermal comfort. Then, reasoning could be applied to identify which room components can be used to change the parameters. For example, an electric heater can be one of the viable solutions because it only increases air temperature. Based on the availability of the room components, the system could suggest non-power-consuming components to reduce electric energy.

Reasoning: In Figure 6, the grey area represents Bob’s comfort range, and only Bob feels comfortable in the gym. Outdoor air-speed (2m/s) is faster than indoor air-speed (0.3m/s). Additionally, outdoor air quality is good, and opening windows (non-power-consuming components) can be a better option for increasing air speed than turning on the air-conditioner (power-consuming component) in terms of energy use. If the window is opened, the air-speed would increase. For example, if the air-speed becomes 1.6m/s, all the occupants can feel comfortable, as shown in Figure 6.

Figure 6: Psychrometric Chart before Opening Windows (left, air-speed: 0.3m/s) and after Opening Windows (right, air-speed: 1.6m/s)

6. Question: In a room, only one occupant sits on a chair. Is this occupant feel comfortable? The occupant has a preferred temperature range of 72°F to 80°F and a preferred humidity range of 28% to 40%. The room temperature is 75°F and the relative humidity is 55%.

Sample Answer from Ontology: The occupant doesn’t feel comfortable.

Terms used from Ontology: air temperature, relative humidity, occupant profile, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, and real-time sensor values such as air temperature, etc., (2) Loading an occupant’s comfort ranges, (3) Reasoning to check whose comfort ranges the current measurements lie within, (4) Showing who is currently comfortable

Usage scenario covered: Three occupants in an office. They have various comfort ranges for different environmental factors, and the system must reason to find out who is comfortable.

Description/Description + Ontology Usage: Based on users’ input data, the system would load BIM/sensor data, as well as the occupants’ upper and lower comfort bounds. The ontology would be leveraged to find out the occupant is comfortable by reasoning about whose comfort ranges the current measurements fall within. For example, the current air temperature and air speed are within the preferred temperature and air speed ranges of the occupant, but not relative humidity. Then, reasoning could be applied to identify if the occupant is comfortable.

Reasoning: The current air temperature is within the occupant’s comfort ranges, but relative humidity is higher than the preferred upper bound. Therefore, the occupant doesn’t feel comfortable in the room.

7. Question: In a small office space with three occupants, who are currently comfortable? Occupant 1 has a preferred temperature range of 64°F to 68°F, prefers lower humidity (25% to 35%), and enjoys a light breeze (1 m/s to 2 m/s). Occupant 2 has a preferred temperature range of 70°F to 75°F, is comfortable in varied humidity (30% to 40%), and likes a light to moderate breeze (1 m/s to 3 m/s). Occupant 3 has a preferred temperature range of 68°F to 74°F, is comfortable in most humidity settings (30% to 50%), and prefers no breeze (0 m/s to 1 m/s). The office temperature is 70°F, the relative humidity is 30%, and the air speed is 2 m/s.

Sample Answer from Ontology: Occupant 2 is comfortable.

Terms used from Ontology: air temperature, relative humidity, air speed, occupant profile, occupant

Semantic Processes Involved: (1) In the knowledge graph, loading BIM database, and real-time sensor values including air temperature, relative humidity, airflow, air quality, etc., (2) Loading an occupant’s comfort ranges, (3) Reasoning to check whose comfort ranges the current measurements lie within, (4) Showing who is currently comfortable

Usage scenario covered: Three occupants in an office. They have various comfort ranges for different environmental factors, and the system must reason to find out who is comfortable.

Description/Description + Ontology Usage: Based on users’ input data, the system would load BIM/weather/sensor data, as well as the occupants’ upper and lower comfort bounds. The ontology would be leveraged to find out who is comfortable by reasoning about whose comfort ranges the current measurements fall within. For example, the current temperature is within the preferred temperature ranges of Occupants 2 and 3, but not Occupant 1. Then, reasoning could be applied to identify if all of the current indoor comfort parameters are within any occupants’ comfort ranges. In this example, only Occupant 2 has comfort ranges that encompass the current environmental parameters. Based on the aforementioned ranges for each occupant, the system can predict that only Occupant 2 is comfortable.

Reasoning: The current air speed and humidity are within Occupant 1’s corresponding comfort ranges, but the temperature is above their preferred upper bound. The temperature, air speed, and humidity are all within Occupant 2’s corresponding comfort ranges. The current humidity and temperature are within Occupant 3’s comfort ranges, but the air speed is above their preferred upper bound. So, Occupants 1 and 3 are not comfortable, but Occupant 2 is comfortable.

VIII. Resources

Knowledge Bases, Repositories, or other Data Sources

External Ontologies, Vocabularies, or other Model Services (partial)

IX. References

[1] US Energy Information Administration. “U.S. energy consumption by source and sector, 2021”, available at https://www.eia.gov/totalenergy/data/monthly/pdf/flow/total-energy-spaghettichart-2021.pdf

[2]**US Energy Information Administration. “Quadrennial Technology Review 2015”, available at https://www.energy.gov/sites/prod/files/2017/03/f34/qtr-2015-chapter5.pdf

[3] ASHRAE Terminology. “indoor environment quality (IEQ)”, available at https://xp20.ashrae.org/terminology/index.php?term=indoor%20environment%20quality%20(IEQ)

[4] Luo, Maohui, Zhe Wang, Kevin Ke, Bin Cao, Yongchao Zhai, and Xiang Zhou. "Human metabolic rate and thermal comfort in buildings: The problem and challenge." Building and Environment 131 (2018): 44-52.

[5] Hasson, Rebecca E., Cheryl A. Howe, Bryce L. Jones, and Patty S. Freedson. "Accuracy of four resting metabolic rate prediction equations: effects of sex, body mass index, age, and race/ethnicity." Journal of Science and Medicine in Sport 14, no. 4 (2011): 344-351.

[6] Tartarini, F., Schiavon, S., Cheung, T., Hoyt, T., (2020). “CBE Thermal Comfort Tool: online tool for thermal comfort calculations and visualizations”. SoftwareX 12, 100563.

[7] International Organization for Standardization. (2005). “Ergonomics of the thermal environment — Analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria (ISO 7730)”. available at https://www.sis.se/api/document/preview/907006/

[8] Hasan, M. H., Alsaleem, F. M., & Rafaie, M. (2016). “Sensitivity analysis for the PMV thermal comfort model and the use of wearable devices to enhance its accuracy”. International High Performance Buildings Conference. Paper 200.

[9] US Energy Information Administration. “Technical Assistance Document for the Reporting of Daily Air Quality – the Air Quality Index (AQI)”, available at https://www.airnow.gov/sites/default/files/2020-05/aqi-technical-assistance-document-sept2018.pdf

[10] Standard, A. S. H. R. A. E. (2017). Standard 55–2017 thermal environmental conditions for human occupancy. Ashrae: Atlanta, GA, USA., available at https://hogiaphat.vn/upload/docs/ASHRAE55-version2017.pdf

Appendix A. Equation of the Predicted Mean Vote (PMV) [7]

The PMV is an index that predicts the mean value of the votes of a large group of persons on the 7-point thermal sensation scale (see Table 1), based on the heat balance of the human body. Thermal balance is obtained when the internal heat production in the body is equal to the loss of heat to the environment.

Table 1: Seven-point thermal sensation scale

                         (1)

                                (2)

                                        (3)

                                                (4)

where

         is the metabolic rate, in watts per square meter (W/m2);

         is the effective mechanical power, in watts per square meter (W/m2);

         is the clothing insulation, in square metres kelvin per watt (m2 ⋅ K/W);

         is the clothing surface area factor;

         is the air temperature, in degrees Celsius (°C);

         is the mean radiant temperature, in degrees Celsius (°C);

         is the relative air velocity, in meters per second (m/s);

         is the water vapor partial pressure, in pascals (Pa);

         is the convective heat transfer coefficient, in watts per square meter kelvin [W/(m2 ⋅ K)];

         is the clothing surface temperature, in degrees Celsius (°C).

Appendix B. Metabolism Estimation based on Age, Sex, Weight, and Height [8]

                                        (5)

where

         is the mass of the body (in kilograms);

         is the height of the body in cm;

         is the age in years;

         is a factor relating to sex, .

        (6)

where

         is the physical activity level

                ;

                (7)

where

         is the activity intensity;

         is the activity duration

                                                                                        (8)

Appendix C. Metabolic Rates for Typical Tasks [10]

Appendix D. Clothing Insulation  Values for Typical Ensembles [10]