A tracking cooling fan using geofence and camera-based indoor localization
Introduction
In warm climates, compressor-based cooling is the main contributor to the energy consumption in buildings. In the United States, compressor-based cooling accounts 13% and 14% of primary energy consumption in commercial and residential buildings respectively [7]. In tropics such as Singapore, the electricity consumed by air conditioning comprises up to 50% of total electricity usage by buildings [22]. Timing and quantity of energy use associated with cooling have large impacts on cost, greenhouse gas emission, peak load of electricity use and reliability of electrical grid.
Raising cooling setpoint of air conditioning system can both bring financial benefits [30], [27], [11], [26], [8] and reduce negative impact on the environment [21] but it may run the risk of sacrificing the comfort level felt by occupants, directly influencing their health, well-being, and productivity.
The challenge can be overcome through elevated air speed generated by electric fans which is a cost-effective and energy-efficient cooling method. Unlike compressor-based cooling systems which lower the air temperature and humidity, electric fans increase the air movement around people. Air movement has significant cooling effect and increases the acceptable range of indoor temperatures [18], [37], [29]. It can be used in conjunction with the air conditioning system.
Most of cooling fans are not connected to the building management system (BMS) and the few that are connected operate based on maximum occupancy assumption, fixed air speed and schedules. Buildings lack intelligent reasoning to customize the fan operation to meet the needs of occupants resulting in wasted energy and sub-optimal thermal comfort. This could be due to the following points:
- (a)
Fans do not know the positions of people. This may cause inconvenience in the cases that the occupant is not willing or does not have the chance to redirect the fan or manually adjust the fan speed, especially when an occupant moves along his workspace or in the building (Fig. 1). Hence both the direction at which the fan blows air and the fan speed setting should be changed accordingly.
- (b)
There is lack of real-time input of occupancy information to the BMS, such as the number and positions of occupants. It is difficult to track the positions of an occupant in a dynamic indoor environment where the layout and the occupancy keep changing.
- (c)
Traditionally, air speed is not measured due to the high cost of omnidirectional hot-wire anemometers.
For the points mentioned above, (a) and (b) refer to the issue of occupancy detection and tracking for fans operation. Point (c) is about the air speed measurement for thermal comfort. The aim of this paper is to develop a tracking cooling fan system to address the problems stated above.
Occupancy information plays an important role in building cooling and fan operation. For the heating, ventilation and air condition (HVAC) system, Daikin employs “intelligent eye” infrared sensor [5] to detect human movement but it is no more than ascertaining whether or not there are people in the room. Mitsubishi utilizes eight “move-eye” infrared sensors [17] in their air conditioning system for human and environment detection by acquiring thermographic data. However, using the infrared sensor may not be accurate on measuring the skin temperature if the sensor is far from the person. Additionally, it is household-oriented and may not be suitable for personal cooling in the office environment. Previously developed smart fans [35], [43] are able to direct the air towards people based on the occupant detection but adjustment of fan speed is not considered. Other smart systems [15], [3] determine the fan speed setting by temperature or humidity but the occupant-fan distance is neglected. Since the air speed attenuates during the propagation, the distance has great influence on the thermal sensation on people.
A number of developed human detection and tracking solutions can be found in the literature, including PIR (passive infrared) sensor [6], [9], [13], RFID (radio-frequency identification) [23], [42], [14] and CO2 sensor [34], [32]. While each of these methods has advantages, it also has its own limitations. For example, PIR sensors based approaches are compromised by other heat sources with strong radiation (e.g., the direct or reflected sun radiation) and can only provide binary information indicating the presence/absence of a person [14]. Most RFID-based systems rely on proximity detection of mobile readers by tags which could be expensive and the density of reference tags is crucial and hard to determine [24]. The performance of CO2-related solutions highly depends on the calibration and commissioning of measurement instrumentations [34].
Custom-digitized geofence has emerged as a significant area of interest for location-based service (LBS) over the last decade. A geofence is defined as a virtual perimeter for a real-world geographic area and LBS can be delivered when a target enters or exits a geofence. Ref. [25] presents fundamental concepts of geofencing and its applications in the transport and logistics sector. Ref. [19] discusses geofence-based notification as general-purpose service from commercial promotion to tourism, traffic information, public service and safety. Ref. [20] employs network proximity rules as geo-based boundaries to effectively deploy indoor location based services and provide significant energy saving for mobile devices compared with the traditional methods.
In this paper, a tracking cooling fan using geofence and camera-based indoor localization is proposed for thermal comfort and energy savings. The cooling area of the electric fan is divided into several sub-area bounded by predefined geofences. By detecting the occupant enter which geofenced sub-area, the direction of air flow is determined. Compared to infrared radiation based PIR sensor or proximity detection based RFID, camera-based vision analysis technique [4], [12], [10] for indoor tracking is able to provide highly accurate position estimates with the resolution being from 0.01 cm to 1 cm [16]. The approach brings convenience for the occupants since there is no need for them to carry extra devices. The proposed indoor tracking system identifies the tracked targets from the images taken by the camera fixed in the tracking area and the captured images are then checked against a pre-calibrated database to give position estimates. The system estimates the position of the occupant, determines the direction of air flow, and calculates the occupant-fan distance. The input power of the geofenced fan is then automatically adjusted by a calibrated mapping algorithm to generate the desired air speed.
The potential of comfort improvement and energy saving by elevated air speed is often ignored due to the strict limits set by thermal comfort standards such as ASHRAE 55-1992/2004 and Singapore Standard SS 553:2009 [33]. Additionally, the most commonly used thermal comfort assessment tool, the PMV (Predicted Mean Vote) index, is not sufficiently accurate in elevated air movement conditions [2]. In recent years, studies of occupied buildings provide consistent evidence that large percentages of occupants prefer more air movement than what they currently have in conditions perceived as warm, thermally neutral and even slightly cool [2]. According to the latest ASHRAE Standard 55-2013 [1], air speed could have no limits or up to 0.8 m/s depending on whether or not the occupant control of air speed is available. In the proposed system, the electric fan is pre-calibrated to obtain the relationship between air speed and fan input power by measuring air speed at different target-fan distances before real-time implementation. The effect of elevated air speed on the sensation of thermal comfort can be assessed through the PMV-SET (Predicted Mean Vote – Standard Effective Temperature) thermal comfort model. In this paper, the PMV-SET model is also used to determine the desired air speed.
Section snippets
Tracking cooling fan system architecture
The flow chart of the proposed system is shown in Fig. 2. The system engine first estimates the position of the occupant and the fan will be automatically switched on/off depending on whether or not the occupant enters the cooling area bounded by predefined geofences. Based on the estimated position , the system engine then calculates the slope of the straight line crossing and the position of the fan located at the original point to direct air flow, and also the
Localization results
The localization results are summarized in Table 3. As shown in Table 3, the estimated coordinates of the five tested positions are given in the 2nd column. Once this information is available, the slope of the straight line crossing the original point and estimated position, and the occupant-fan distance can be calculated as shown in the 3rd and the 5th columns respectively. The former metric is used to decide at which sub-area the fan should blow air while the latter one helps to select the
Automatic mode and occupant-control mode
The proposed system can actually be set in two different operation modes: the automatic mode and the occupant-control mode. The first mode refers to the methodology described in the paper and it strictly follows the ASHRAE 55-2013 standard that the air speed should be limited below 0.8 m/s. Thus the fan can be only used up to the 12th level of speed setting (Table A1 in Supplementary Information). However, this is not applicable for the occupants who want to take control over the thermal
Conclusion
In this paper, a tracking cooling fan using geofence and camera-based indoor localization is proposed for thermal comfort improvement and energy savings. The proposed camera-based indoor tracking system is able to accurately locate the positions of the occupant, determine the direction of air flow, and calculate the occupant-fan distance. A calibrated mapping algorithm is also proposed to automatically adjust the fan input power to generate desired air speed determined by the PMV-SET model.
The
Acknowledgement
This research is funded by the Republic of Singapore’s National Research Foundation through a grant to the Berkeley Education Alliance for Research in Singapore (BEARS) for the Singapore-Berkeley Building Efficiency and Sustainability in the Tropics (SinBerBEST) Program. BEARS has been established by the University of California, Berkeley as a center for intellectual excellence in research and education in Singapore.
References (43)
- et al.
Towards a sensor for detecting human presence and characterizing activity
Energy Build.
(2011) - et al.
Building occupancy detection through sensor belief networks
Energy Build.
(2006) - et al.
Revealing occupancy patterns in an office building through the use of occupancy sensor data
Energy Build.
(2013) - et al.
Extending air temperature setpoints: simulated energy savings and design considerations for new and retrofit buildings
Build. Environ.
(2015) - et al.
Location-based human-adaptive air conditioning by measuring physical activity with a non-terminal-based indoor positioning system
Build. Environ.
(2013) - et al.
Measuring and monitoring occupancy with a RFID based system for demand-driven HVAC operations
Autom. Constr.
(2012) - et al.
Novel activity classification and occupancy estimation methods for intelligent HVAC (heating, ventilation and air conditioning) systems
Energy
(2015) - et al.
Energy saving and improved comfort by increased air movement
Energy Build.
(2008) Higher space temperatures and better thermal comfort—a tropical analysis
Energy Build.
(1995)- et al.
Development and In-situ validation of a multi-zone demand-controlled ventilation strategy using a limited number of sensors
Build. Environ.
(2012)
In-situ implementation and validation of a CO 2-based adaptive demand-controlled ventilation strategy in a multi-zone office building
Build. Environ.
Cooling efficiency of a brushless direct current stand fan
Build. Environ.
Comfort under personally controlled air movement in warm and humid environments
Build. Environ.
Effect of local exposure on human responses
Build. Environ.
Thermal sensation and comfort models for non-uniform and transient environments: part I: local sensation of individual body parts
Build. Environ.
Thermal sensation and comfort models for non-uniform and transient environments, part II: local comfort of individual body parts
Build. Environ.
Thermal sensation and comfort models for non-uniform and transient environments, part III: whole-body sensation and comfort
Build. Environ.
A comprehensive multi-factor analysis on RFID localization capability
Adv. Eng. Inf.
ANSI/ASHRAE Standard 55-2013, Thermal Environmental Conditions for Human Occupancy
Moving air for comfort
ASHRAE J.
Cited by (31)
Computer vision to advance the sensing and control of built environment towards occupant-centric sustainable development: A critical review
2024, Renewable and Sustainable Energy ReviewsHuman-building interaction for indoor environmental control: Evolution of technology and future prospects
2023, Automation in ConstructionA fusion framework for vision-based indoor occupancy estimation
2022, Building and EnvironmentApplication of vision-based occupancy counting method using deep learning and performance analysis
2021, Energy and BuildingsCitation Excerpt :In addition some studies proposed to automatically extract the number of occupants without human intervention and delete the saved images immediately [46–48]. In addition, video encryption and security was enhanced in some studies [39,44,49]. In another study, occupancy counting was performed by tracking occupants entering and leaving the entrance instead of photographing indoors [50–52].
Review of vision-based occupant information sensing systems for occupant-centric control
2021, Building and EnvironmentCitation Excerpt :Some researchers used MATLAB [37] or Python [52] to implement complex control algorithms in simulations. Several field measurement studies have been conducted for short periods of time in controlled environments, such as laboratories and chambers [60,61,63]. Only three studies have performed long-term field measurements in uncontrolled environments [23,50,51].
User-centered environmental control: a review of current findings on personal conditioning systems and personal comfort models
2020, Energy and BuildingsCitation Excerpt :Although air temperature and relative humidity are the most frequently used variables in monitoring during operation, as shown in Table 1, it is necessary to predict personal comfort personal variables, which are captured by sensors and new proposed technological systems. In the case of the study by Liu et al [86], occupancy sensors used by [67,83,84] would be insufficient to detail the position of users for fans rotation automatic adjustment. Therefore, the authors propose a tracking system with video georeferencing.