Wireless energy efficient occupancy-monitoring system for smart buildings

Abstract

Rationalizing energy consumption in smart buildings is considered in this paper, and a wireless monitoring system based on Passive Infrared sensors (PIRs) is proposed. The proposed system is pervasive and can be integrated in existing buildings without any complicated wiring or setting. Realistic constraints are considered for this purpose such as sensing-hole, battery limitation, user comfort, etc. To ensure maximum coverage in presence of holes, the optimal placement of PIRs is formulated as a mixed integer linear programming optimization problem (MILP). Experimentations have been carried out to quantify the effects of the holes on the detection accuracy and to demonstrate the impact of the optimal PIRs placement on energy consumption. To facilitate installation and integration without complicated settings, notably in existing buildings, the system is designed to be battery operated. Therefore, energy efficiency will not be limited to optimize energy consumption in buildings, but also to optimize consumption in the components of the system (sensors and actuators). Duty cycling is inevitable to extend the network lifetime of such components, but the setting of this cycle yields a trade-off in optimizing the energy consumption i) at the building level, vs., ii) that consumed by sensors and actuators. Reducing energy consumption (duty cycle) of sensors/actuators will delay non-occupancy detections and thus will increase the building energy wastage, and vice-versa. Duty cycling the radios is dealt with and modeled as a cooperative game, which allows to derive a Nash Bargaining as the optimal balancing cycle. The proposed approach is analytically investigated using realistic parameters of the existing hardware and users’ comfort. The results demonstrate that the system can survive for more than 6 years without battery replacement.

Introduction

Energy consumption in residential and commercial buildings has increased dramatically worldwide in the last decade as an inevitable consequence of the proliferation of electronic and consumer appliances, as well as the constant economic and population growth notably in urban areas. The later cover $2\text{\%}$ of the earth’s surface but are responsible of $78\text{\%}$ world’s energy consumption and $60\text{\%}$ of CO${}_2$ emission [1]. Buildings are the cause of $40\text{\%}$ of the total energy consumption in the US and the EU [2]. Similar figures are reported worldwide, which makes the building sector the main source of energy consumption and expected to remain as such in the next decades [3]. Moreover, it has been estimated that as much as $30$-to-$50$% of the building’s energy is wasted due to misuse and non-optimal management [4]. This motivates research efforts towards improving and modernizing Building Energy Management Systems (BEMS) in the last few years. Such systems should have attributes from all facets of building to control and manage functions such as Heating, Ventilation and Air-Conditioning (HVAC), lighting, fire alarm system, etc. An efficient design of BEMS allows to achieve smooth buildings’ operations and maximize energy saving while preserving users’ comfort. Recent advances in ubiquitous wireless communications and sensing technologies have promoted the deployment of Wireless Sensor Networks (WSNs) in many application areas, including BEMS. Battery operated WSNs add flexibility to BEMS and allow deployment without hard or intrusive installations. This is particularly motivating in old buildings that does not incorporate any intelligent BEMS system. In such buildings, the installation of a BEMS system maybe impossible and very expensive due the lack of the basic standards in the building structure to support such a sophisticated installation and where the modification of the existing structures is usually undesirable. Further, the solution can also be used in modern buildings to reduce cost and facilitate maintenance compared to existing wired BEMS. For instance, when modifying space partitions to make new or extend the size of the offices, additional wiring is required to connect to the existing BEMS. By using our solution, both the installation cost and times are minimized. However, factors such as the use of low cost components (e.g., infra-red sensor for occupancy detection) and supplying the system with batteries raise some design challenges on the system accuracy, reliability and sustainability. Detecting user occupancy in buildings is a fundamental step for reducing wastage of energy and improving users’ comfort. In fact, most BEMS in old buildings use a set of predefined actuation schedules for managing electrical appliances, such as HVAC and lights. These schedules have a coarse-grained time dependability that is generally related to static issues such seasons, days of the week, etc. However, by dynamically detecting vacant places, more optimized context-aware schedules can be implemented to shorten the actuation durations without compromising the user’s convenience. Many of the solutions proposed for tracking the presence of occupants in buildings are based on the use of passive infrared (PIR) sensors [5], [6], [7], [8], [9]. These sensors are made from inexpensive pyroelectric materials that react to the change of infrared emissions in the environment, which helps in capturing the presence of humans in a specific space. The low cost and low energy consumption of such sensors enable their large use in battery-operated wireless systems. Further, they do not affect the privacy of people, contrary to other sensors such as cameras and microphones. This privacy preservation make such sensors appropriate for monitoring private spaces such as offices, meeting rooms, homes, etc.

However, a major drawback of PIR sensors is the false negatives (non-detection) in some situations. The first reason behind this shortcoming is that these sensors are only capable of detecting motion, but not static bodies. Whilst this does not represent any problem in many premises of buildings where people are moving, such as corridors and near to the doors, it prevents accurate monitoring in places such as offices where workers tend to stay immobile for relatively long periods. To tackle this problem, some solutions have been proposed in the literature that complement the PIRs with information provided by additional sensors. For example, the occupancy detection system proposed by Agarwal et al. [5] is enhanced with a magnetic reed switch sensor that tracks the open/close events of an office door. This information is matched with the output of the PIRs. ThermoSense [8] uses, in addition to a PIR sensor, a thermal sensor array that is able to measure temperatures of a $2.5\mathrm{m}\times 2.5\mathrm{m}$ area discretized as a 8 × 8 grid. Alternatively, some other solutions use other sensing techniques, such as [10]. The second problem is that the sensing area of a typical PIR module is not a contiguous volume, but it includes spaces where changes of infrared emissions are not captured by the sensor. We refer to these uncovered spaces by, sensing-holes. The dimensions of these sensing-holes become larger as the distance separating the sensor to the detection zone increases. For instance, with state-of-the-art PIRs, it reaches the scale of a human body movements at a distance of $2\mathrm{m}$ to $3\mathrm{m}$, which represents typical height of ceiling at offices where PIRs are usually installed. Consequently, a PIR cannot detect a person within the sensing-hole even when he performs small movements (e.g., in the office scenario, moving his arms, his head, rotating the chair when sitting, etc.). While it seems infeasible to detect static body only with PIRs (the first problem), it is possible to tackle the second one by investigating the sensing-holes and their impact, and using optimal deployment of PIRs to eliminate/minimize such holes.

In addition to the false negatives, lifetime of battery operated sensor has always been an issue in real deployments. Frequently replacing the batteries after installation is impractical and makes the solution unattractive. The trend in many applications is to use energy harvesting technologies to supply sensor motes. Solar energy is currently the most effective source given its high efficiency as compared to other technologies such as wireless charging. However, this cannot be used for indoor deployment, which features the application considered in this work. A possible alternative in the future will be wireless recharging, but it does not seem possible in the short or mid-term horizon to achieve reasonable charging efficiency with this technology. This makes optimal energy management of batteries the only remaining option. Given that the radio consumes the largest amount of a node’s battery [11], the only way to extend the lifetime is to duty-cycle the radio component and repeatedly switching it between active and sleep modes. In active mode, a node can receive and transmit packets, while in the sleep mode, it completely turns off its radio to save energy. Using low cycles with high period of sleep mode trivially allows to extend the battery lifetime but may delay reporting of detections to the control system and/or the appropriate actuator (also called switch-mote in simplified settings), which have undesired effect on the user comfort and/or the BEMS. Whereas high cycles reduces this problem but at the cost of reducing the batteries’ lifetime.

In this paper, we tackle all the above mentioned problems and propose an efficient yet low cost occupancy detection system for energy saving in buildings. For occupancy monitoring, the system uses only PIR sensors. This facilitates installation in existing buildings and even in buildings that does not use any BEMS (e.g., in developing countries). The main contribution of this paper are summarized in the following.

• We propose a solution to the occupancy detection accuracy using low cost PIR sensors and formulate the problem of placing minimal number of PIR that maximizes the coverage of the monitoring area with mixed integer linear programming. Realistic features such as sensing holes are considered in the model. A short version of the solution has been already published in [12].

• We consider maximizing system lifetime by duty-cycling different components of the system with the use of PIRs to trigger their wakeup and ensuring user comfort. Without loss of generality, we consider a simple setting in offices with PIRs/light sensors that monitor the occupancy and day light, respectively, and switch actuator that react upon detections. We determine the optimal duty-cycle period that tradeoff the PIR sensor motes lifetime with the switch by defining a cooperative bargaining game model between the two motes. This model allows to derive the Nash Bargaining point as the optimal balancing cycle.

• Analytical and experimental results are provided to validate the proposed approach and empirically demonstrate the efficiency of the solution in saving electrical energy while ensuring user comfort.

The remaining of the paper is organized as follows. Section 2 states some existing electrical energy management systems and sensor based solutions. Section 3 presents the proposed solutions for the occupancy monitoring system, starting with a general overview of the proposed system in Section 3.1, followed by the occupancy detection in Section 3.2, and then the duty-cycling solution, in Section 3.4, which has been proposed for extending the system lifetime. The proposed game theory-based model of duty-cycle balancing with the different experiments results are presented in Section 4. Finally, Section 5 draws the conclusions.