Smart System for Social Housing Monitoring

Smart System for Social Housing Monitoring Ammar ALJER, Marine LORIOT, Isam SHAHROUR Afif BENYAHYA Laboratory of Civil Engineering and geo-Environment Lille University, 59650, Villeneuve d’Ascq, France Lille Métropole Habitat (LMH) 425 Boulevard Gambetta, 59200 Tourcoing, France Abstract—This paper presents the design, realization and use of an innovative system for social housing monitoring. The specifications of the system were determined through concertation with social housing tenants as well as technical and administrative staffs of the social housing manager. The system aims at monitoring the fluids consumption as well as indoor comfort parameters using robust, wireless, low-energy consumption, low-cost and friendly interface system. The paper presents this system and its use in a demonstration apartment of social housing. Keywords— Smart System; Sensor; Multi-sensor; Social; house; Indoor; Innovation; Low-cost I. INTRODUCTION Social housing plays a major role in France in hosting inhabitants with low income. With 4.6 million units, the social housing stock accounts for 17% of the total housing. Since the major part of this stock is old, it suffers from poor quality as well as high running expenses and degraded life environment. With the increase in concern for the quality of life and reduction of the energy consumption and greenhouse gas emission, public authority and social housing managers are interested by the development of innovation to meet the increasing challenges in the sector of social housing. This paper presents the development of an innovative system, which was designed to monitor social housing, with the objective to understand the indoor condition, equipment functioning and tenants’ behavior in order to establish a basedknowledge strategy to improve the social housing efficiency and quality. This development was carried out within a joint program between Lille University and a social housing manager “Lille Metropole Habitat”. The program included different phases: (i) determination of the monitoring system specifications, (ii) design and construction of this system (iii) its verification in occupied dwells (iv) its use for the analysis of the indoor environment and tenants’ behavior and (v) its use for the improvement of the social housing efficiency. This paper presents the first results of this project. II. INNOVATIVE SYSTEM FOR SOCIAL HOUSING MONITORING A. Specifications Important work was conducted to reduce the energy consumption and the greenhouse gas emission ([1], [2]). Since the social housing concerns mainly low-income inhabitants who live in small units, but in large collective and old buildings, it was important to start by monitoring the indoor environment to understand the real condition inside these buildings as well as the expenses related to fluids consumption and the behavior of tenants. To establish the specifications of the monitoring system, a concertation was conducted with tenants of the social housing as well as with the technical and administrative staffs to analyze their expectation and demand. The concertation showed a need for an innovative monitoring system, which could help the tenants to follow fluids consumption (water and energy), comfort conditions (temperature, humidity, air quality, lightening, noise) and state of windows and doors (open/closed). The system should store data and allow analysis of historical data. It should also include friendly graphic interface and guarantee tenants’ privacy. The system should also be robust, based on wireless low energy consumption technology and low-cost. The first phase of this work did not include the equipment control. This phase will be investigated after the understanding of the indoor environment, the equipment functioning and tenants’ behavior. Fig. 1 summarizes the specifications of the system as determined after concertation with tenants and the technical and administrative staffs. • Real-time monitoring of water and energy consumption • Real-time monitoring of indoor comfort (temperature, humidity, air quality, lighting, noise..) • Real-time control of the doors and windows (open/closed) • Storage and analysis of historical data • Privacy guarantee • Friendly graphic interface • Ease of installation (Based on wireless low energy technology) • Robust and low-coast. Fig. 1. Specifications of the social housing monitoring system 978-1-5090-6011-5/17/$31.00 ©2017 IEEE B. Design of the monitoring system Since the commercial offers did not meet the specifications summarized in Fig. 1, a new system was designed. It is composed of 3 parts: central unit, wireless sensors and friendly users’ interface (Fig. 2) Cozir CO2 sensor is used for the air quality; it works in the interval 0 to 5 000 ppm with a precision of 3%. Luminosity is tracked using TSL2561 sensor, which works in the interval 0.1 to 40 000 lux, Noise is tracked using audio sensor including a 100Hz to 10kHz electret microphone associated with a 60x amplifier. All the sensors are associated with a PanStamp programmable low-power wireless board (module), especially conceived for Internet Of Things applications, with an Atmega328p micro-controller and a CC1100 RF transceiver. It consumes only 1 uA in sleep mode and 2.5 mA in transmitting mode. It could be programmed with the Arduino Environment. Since the communication protocol is entirely implemented by PanStamp, developers can focus on development of applications of the sensor board. Fig. 2. Architecture of the monitoring system Izar Pulse I, magnetic contact and current transformer sensors are associated with a PanStamp Battery-Board, which includes a card with the Panstamp wireless module, powered by a AA battery. It provides analog and numeric inputs to get sensors signal and to transmit it with RF to the central unit. The central unit ensures the management of the monitoring system. It communicates with sensors using radio frequency (RF) protocol It hosts free and open software for data storage, analysis and display. It is composed of a Raspberry Pi, a small computer without screen or keyboard, which uses the free and open-source Linux operating system. Storage is ensured by a SD card. Temperature, humidity, luminosity, CO2 and audio sensors are integrated into a single electronic card. The card is powered by a AA battery; it uses the PanStamp wireless communication module. Since the energy consumption is mainly due to the activation of the RF communication, a single card with multiple sensors reduces energy consumption. The central unit offers the possibility to create a local Wi-Fi network, which allows tenants to access to data and information stored in the central unit. It does not require internet connection. D. Communication protocol The PanStamp Wireless module uses the open-source Simple Wireless Abstract Protocol (SWAP). It can use the 868/905/915/918 MHz free Industrial, Scientific and Medical (ISM) frequency bands. Our system uses 868 MHz frequency. It works within an open area of around 200 meters. Wireless sensors are used for tracking the parameters summarized in Fig. 1. They are connected to the central unit via radio frequency protocol at optimized time intervals. To reduce the energy consumption and the monitoring system cost, we have developed a multi-parameters smart card, that can monitor several parameters and send data using only one communication system. A web friendly interface was designed to enable users to access easily and friendly to all the information concerning the indoor environment and consumptions. C. Sensors Water Consumption is measured using Izar Pulse transmitter associated with an induction sensor. Current transformers are used to measure electricity consumptions. Magnetic contact detectors are used for the determination of the doors and windows state (open or and closed). Temperature and relative humidity are measured using SI7021 sensors. The temperature is measured in the temperature interval -10 to 85°C with 0.4% precision, while the relative humidity is measured in the interval 0 to 80% with 3% precision. The SWAP software stack ensures listening and parsing incoming SWAP packets then transmitting them or responding to their queries or command, management of registers, sending updated data and managing power. For each sensor, the configuration parameters and data are stored in a register, which is the SWAP unit with a unique identifier. The frequency at which each register is updated and sent with RF can be chosen or the update is triggered by event. It is also possible to include different sensors data in one register as its size can be up to 55 bytes. Configuration parameters as security options and device address are stored in specific registers and the developer can use up to 245 custom registers. A SWAP frame includes the addresses of the destination and source devices, hop counter, security options, security nonce, function of the packet, address and identifier of the register and finally the register value which is the payload. The Raspberry Pi is equipped with a PanStamp wireless board to follow the incoming SWAP packets. E. Data Management The design of the Database permits to easily match the physical addresses of the sensors with their semantic values and to reference the dynamic values regarding the time [3][4]. MySQL, the free and open-source RDBMS, is used for data management. During the installation of the system, the sensor boards are fixed in the different parts of the habitat. Then the unique address of the board is attached in the database to its emplacement in the habitat. Received data are classified and registered in the database regarding their location. A Java application follows the incoming SWAP packets received by the PanStamp wireless module, plugged on the Raspberry Pi. The application checks for each frame that the source device is registered in the database. Then, each type of value of this board is stored in the history database table that contains all sensors values of the apartment. The Java application conducts treatment to present data in understandable units to tenants such as cost and quality of indoor environment. G. Maintenance Alerts are generated in case of any fault in the system, which could result from power shortage or connection. Alerts are sent to users, who can operate some actions such as change the battery or re-start the system. More serious faults require the intervention of the technical staff. Since this system is used for research, regular visits are conducted at 3-months interval to check the system, to discuss with users about its use and to learn from their feedback. III. USE IN A DEMONSTRATION APARTMENT The monitoring system was installed in a demonstration social housing apartment, which is composed of 2 bedrooms, kitchen and living room. Fig. 4 shows the monitoring system used in this apartment. Humidity and temperature sensors were installed in the bedrooms and the living room. Air quality sensor was installed in the living room; electrical consumption sensor was installed at the electrical supply counter and contact sensors were installed at the windows and doors. To ensure tenants’ privacy, data are stored in a local database, which is accessible only by the tenants, who have the possibility to provide data accesses to other users. F. Visualization The central unit contains an Apache web server, which permits users to access via a friendly web interface to real-time and historical data using graphic interface (Fig. 3). The web interface is implemented in HTML, CSS and JavaScript with Bootstrap and Highcharts libraries to allow the design of interactive charts. PHP is used to communicate with the database to get the sensors’ values. The web server is accessible via a local Wi-Fi network using smartphones, tablets and Smart TV. Fig. 4. Monitoring of the demonstration apartment Fig. 5 displays the variations of the temperature in the living room and the bedroom during 2 days in February. The temperature in the living room varies around 20°, while the temperature in the bedroom varies around 19°. This figure shows that the regulation of the heating system in the apartment works well and ensures the regulation requirement (temperature around 19°) with a small decrease in the temperature during sleeping hours. However, this decrease is very small and could be augmented by 2° for energy saving. Fig. 3. Users interface Fig. 5. Variations of the temperature (red : Bedroom, bleu Living room) Fig. 6 displays the variation of the relative humidity in the living room and bedroom. It varies between 28% and 37%. The humidity in the bedroom is higher than that in the living room; this difference could result from a higher ventilation of the living room. The humidity is also high in the morning and in the evening. These high values could be related to the human presence as well as indoor activities such as cooking. IV. CONCLUSION This paper presented the design, fabrication and use of an innovative system for monitoring fluids consumption and comfort parameters in social housing. The specifications of the system were determined through concertation with social housing tenants as well as technical and administrative staffs. The system is based on the use of a central unit (Raspberry Pi), which tracks and controls the indoor environment (temperature, humidity, air quality, lighting, noise..), the water and energy consumption as well as the state of doors and windows (open/closed). The system uses a friendly interface, which allows users to follow real-time data and to access to historical data enhanced by information concerning the quality of the indoor environment and the expenses. The use of the monitoring system in a demonstration social housing apartment confirmed its good performances and robustness. Now, it is used to monitor 15 occupied social housing apartments. REFERENCES Fig. 6. Variations of the relative humidity (red : Bedroom, bleu Living room) Fig. 7 shows the influence of the opening of the window of the living room on both the temperature and humidity. It shows that the opening of the window induces a rapid increase in the humidity and a small decrease in the temperature. [1] [2] [3] [4] Fig. 7. Impact of the opening of the window of the living room on the room temperature (bleu) and humidity (green) G. Welson, T. Hargreaves and R. Hauxwell-Baldwin, “Benefits and risks of smart home technologies” Energy Policy, vol. 103, pp. 72–83, April 2017. T. K. L. Hui and R. S. Sherratt, “Major requirements for building Smart Homes in Smart Cities based on Internet of Things technologies”, Future Generation Computer Systems, In Press Available online; 1 November 2016 C. Gutierrez, S. Servigne and R. Laurini, “ Towards Real Time Metadata for Network-Based Geographic Databases” 5th International Symposium of Spatial Data Quality, Enschede, IEEE Xplore,13-15 June 2007. M. Arenas-Martinez, P. Roth and A. Zeier, “A Comparative Study of Data Storage and Processing Architectures for the Smart Grid” IEEE International Conference on Smart Grid Communications”, 4-6 October 2010