Open-Hardware Meets Open Software for Environmental Monitoring

Transcription

1 International Environmental Modelling and Software Society (iemss) 7th Intl. Congress on Env. Modelling and Software, San Diego, CA, USA, Daniel P. Ames, Nigel W.T. Quinn and Andrea E. Rizzoli (Eds.) Open-Hardware Meets Open Software for Environmental Monitoring Jeffrey M. Sadler, Daniel P. Ames, Rohit Khattar Brigham Young University ) Abstract: Environmental data are critical to understanding environmental phenomena. This paper describes the design and development of a low-cost environmental monitoring system including sensor, data logger, and transmitter using several low-cost, open-source, mass-produced components. The system is connected to the open source Hydrologic Information System (HIS) software created by the Consortium of Universities for the Advancement of Hydrologic Sciences Inc. (CUAHSI) through a web service interface to the HIS HydroServer Lite. It is expected that this connection will enable the ingestion into the CUAHI HIS of near-real-time data collected by our low-cost device or similar devices through HydroServer. By publishing data in this way, it becomes discoverable through the GIS-based CUAHSI tools, HydroDesktop and HydroShare. Additionally, because it follows WaterML encoding, open hardware data stored in the HIS can be included in international catalogues such as the Global Earth Observation System of System (GEOSS) catalogue. Keywords: Open-source hardware; Sensors; Environmental monitor; Open-source software. 1 INTRODUCTION 1.1 Problem Statement With the development of open-source hardware, the cost of systems used to collect, store, and transmit environmental data has been significantly reduced. As a result of this dramatic reduction in costs, more data than ever before is being harvested and stored in digital format (Barniuk, 2011). However, to maintain the value of collected data, it must be stored and made discoverable, in a structured and standardized way (Borgman, 2012). The purpose of this research is to design, develop, and test a userfriendly, open source hardware sensor linked to open source software tools for sharing hydrologic and climate data on a federated global data network. 1.2 Background In recent years, the advancement of open-source hardware has sparked significant reductions in costs of scientific equipment in many technical areas. For example, the Arduino board, an open-source, programmable microcontroller which is adaptable to both simple and complex tasks (arduino.cc), has been coupled with 3D printers to manufacture custom equipment for experimental laboratories (Pearce, 2012). One such piece of equipment, called a colorimeter (a sensor used to measure chemical concentration), has been produced using open-source hardware, the final cost of which was two orders of magnitude less than its commercial counterpart (Anzlone, Glover, & Pearce, 2013). Open hardware sensors are being used on a global scale in reducing costs of environmental monitoring. For example, an Arduino Pro with a passive infrared sensor was used to monitor latrine usage in Orissa,

2 India. In this research a custom sensor system was produced for approximately $60 USD; a commercial alternative would have cost $400 USD (Clasen et al., 2012). In the Acronet Paradigm, an open hardware project of Acronet Sri ( multi-node sensor systems using open-source parts have been built to measure variables such as snow-depth, soil moisture, and air quality for municipalities in Northern Italy (Fedi, Ferrari, Lima, Pintus, & Versace, 2013). A similar project, SEMAT, is a low-cost sensor system for monitoring marine environments in Australia (Trevathan et al., 2012). Of particular significance in environmental monitoring is the increased spatial and temporal resolution made possible by sensor nodes built with open-source hardware; as each open hardware unit costs significantly less than commercial alternatives, many more nodes can be installed for a given financial investment. The Stroud Research Center, located in Pennsylvania, has taken full advantage of this by building a variety of data acquisition modules with open-source hardware to collect data in the Christina River Basin Critical Zone Observatory. Using Arduino boards and other open-source electronics, the Stroud Research Center has built and configured several custom sensor modules including a radio reporting stream gauge, soil moisture sensor, pressure transducer readout, and respirometer controllers. With sensor and transmission costs reduced by more than an order of magnitude, the Stroud Research Center has been able to obtain data at a greater temporal and spatial resolution (Hicks, Aufenkampe, & Montgomery, 2011). 1.3 Challenges with the current state of the technology Despite much success with open-source electronics in environmental monitoring, there are still improvements to be made in managing the collected data. Open-source sensors dramatically increase the amount of data that can be affordably collected but the value of those data depends on how they are managed and made available for sharing (Piwowar, Day, & Fridsma, 2007). Data management and sharing plans are also emphasized heavily by funding institutions such as NSF (Tenopir et al., 2011). The management and accessibility of open hardware data is an area that can be improved. For example, in the cases of the Acronet and SEMAT sensors there is no plan specified to make the collected data publically available. In the case of the latrine usage sensor in India, the collected data was simply stored on an SD memory card. The Stroud Research Center made their data viewable via a custom interactive website, but no standards for the storage and retrieval of the data were specified. The researchers for the SEMAT project also created a custom data management system. It is noted that in each of these projects, the chosen method for data management and sharing was sufficient and suitable for each particular need. However, this paper describes a more generalized and simpler approach to the management of data collected by low-cost sensors. The simplicity of implementation and the accessibility of the data are two improvements which this research will address. 1.4 Proposed Solution The proposed solution of this research was to build a low-cost sensor configured to automatically input data into the Consortium of Universities for the Advancement of Hydrologic Sciences Inc. (CUAHSI) Hydrologic Information System (HIS). The CUAHSI HIS is a standards-based system for the storage, sharing, publication, and discovery of hydrologic and other environmental time-series data (Horsburgh et al., 2009). This system as well as the low-cost environmental monitor which was built are described in detail in the methods section below.

3 2 METHODS 2.1 Hardware The objective in building the hardware was to assemble and configure a self-contained module to record environmental data. The module also had to accommodate the automatic ingestion of the data into the CUAHSI HIS. In addition, the ability to assemble and configure the sensor easily (e.g. with minimal soldering) and inexpensively were significant considerations in the methodology.. To coordinate the various components of the system, the Arduino Uno microcontroller was chosen. Though the Arduino Uno has greater power consumption and smaller memory capacity than several alternatives, its price, ease-of-use, and availability of support were factors which led to its selection. Other Arduino boards with larger memory capacity and/or smaller power consumption were more expensive, not as user-friendly, or both. For example, the Arduino Mega 2560 R3 with eight times the RAM available to the Uno cost $65 USD while the Uno was purchased for $30 USD (prices from Conversely, some less expensive boards such as the Arduino Pro, can also be configured to consume much less power than the Uno. However, this is possible, in part, at the sacrifice of the USB port; with the Arduino Pro, the user would need to upload any programs to the Arduino via an FDTI cable, which is much less common and user-friendly than a standard USB cable. As usability was a major objective in this research, this difference was a significant swaying point and led to the using of the Arduino Uno with its USB port instead of the Arduino Pro. The Arduino Uno was programmed using the Arduino IDE which is free and available to download at the Arduino website (arduino.cc). In the Arduino IDE, The user writes a script or a sketch (the term used in the IDE) in a programming language derived from C++. The sketch is then uploaded over the USB port to the Arduino board. The uploaded contents of the sketch are instructions to the Arduino board as per tasks it is to do, the coordination of inputs and outputs etc. In the present case, the sketch coordinates the timestamps, sensor readings, recording of the data to variables and to an SD memory card, the transmission of the data, and the power consumption of the modem. The sketch is, as with everything written for Arduino, open-source. As the focus was on the transmission and automatic storage of environmental data, the type of data collected was kept simple: air temperature and relative humidity. Notwithstanding the simplicity of the parameters recorded, it is assumed that these methods are customizable to a variety of environmental variables. In this case, a DHT22 digital temperature and humidity sensor was used. This was chosen on the basis of cost (~$11) and usability. To interface the sensor with the Arduino, a DHT22 library found on GitHub was used. However, this library is only one of several DHT22 libraries found with a simple web search (Adams, 2012). Besides being stored in the Arduino Uno memory, temperature and humidity values are stored onto a Micro SD card attached to the General Packet Radio Service (GPRS) shield used. This shield is also used to transmit the data over the data layer of the cell phone network. The GPRS shield provided realtime clock functionality, data storage on a Micro SD card, and the transmission of the data and timestamps over the internet via an HTTP GET request. Because the GPRS shield relies on the cell phone network, a cellular plan with basic data capabilities, and a SIM card were needed. Both the plan and the SIM card were purchased from a company called PTEL. The SIM card cost $5 USD while the plan, which was non-contract and lasted two months, cost $10 USD. The purchased plan allowed for data to be sent at a rate of ten cents per megabyte up to the ten dollar limit. This was more than sufficient for the needs of this project ( To create a self-contained unit, a solar powering system was needed. To meet the power consumption needs of the Arduino board and the GPRS module, both a solar panel and a battery were required; the solar panel would harvest energy from the sunlight in the daytime and store that energy in the battery to be used in the night-time or when cloudy. A satisfactory solution was found from a manufacturer called Voltaic ( A 6W panel with an accompanying 4000mAh lithium ion battery were purchased for a total of $88, which constituted the largest portion of the total cost. To house the unit, a waterproof, durable plastic container built by Pelican Cases (pelican-case.com) was purchased for $15. In order to get the solar panel wire into the case with the battery, a hole was

4 drilled into the side of the case. With the sensor wires as well entering the box from the outside through the drilled hole, the temperature and humidity outside the case were measured rather than the parameters inside the case. A summary of the cost of each item as well of the total cost is shown in Table 1. Figure 1 is a photo of the solar panel, battery, Arduino, GPRS Modem, temperature sensor, and weatherproof box. Table 1: Individual Item Costs and Total Cost of Hardware Item Cost Arduino Uno $30 Sim900 GPRS/GSM Board $40 SIM Card $5 Phone Plan (good for two months) $10 8GB MicroSD Card $10 DHT22 Temperature and Humidity Sensor $11 Voltaic 6W Solar Panel and 4000 mah Battery $88 Case $15 Total Cost $209 Figure 1: Arduino, solar panel, battery, sensor, and GPRS modem 2.2 Software The software side of the configuration begins with the transmission of the data via an HTTP GET request. The request can be modified in the Arduino code to submit the GET request to any URL. For our case, the GET request was sent to a webserver hosted at BYU, therefore, the following URL was used: In the request, two parameters were transmitted, info and readings. The info parameter contained metadata information and the readings parameter contained the temperature, humidity, and timestamp values. The HTTP request was sent to a PHP script on the server, parsetest.php. The script parses the data into an SQL query which inserts the metadata and data into a persistent, structured data model called the Observations Data Model (ODM). ODM is the relational model used in the CUASHI HIS to store observation data, and the context for that data (i.e. the metadata) in an SQL database (Horsburgh, et

5 1 7:30:00 17:30:00 12:30:00 6:30:00 15:30:00 2:30:00 10:30:00 19:30:00 4:30:00 1 6:30:00 16:30:00 1 9:30:00 18:30:00 11:30:00 19:30:00 12:30:00 Sadler, Ames, and Khattar / Open-Hardware Meets Open Software for Environmental Monitoring al., 2009). For the purposes of this research, the MySQL version of ODM was used with its accompanying PHP user interface, HydroServer Lite. Because it is built on PHP and MySQL, HydroServer Lite is very inexpensive, with the only expense being basic web hosting. Additionally, its simplistic, web-based user-interface makes it very user-friendly. Notwithstanding its low cost and simplicity, HydroServer Lite provides many useful features including a web map interface showing the locations of data-collection sites, graphical and tabular representations of the collected data, and download functionality(conner, Ames, & Gill, 2013; Kadlec & Ames, 2012). The complete workflow, as described above, is shown graphically in Figure 2. Figure 2: Workflow of Data from Sensor to User Aside from it being free and simple to use, HydroServer Lite becomes a gateway to the broader CUAHSI HIS. HIS provides a well-defined system and several software tools to make data easily discoverable and downloadable through standards-compliant WaterML web services called WaterOneFlow services. Included in the WaterOneFlow services are methods such as GetValues and GetSiteInfo. Each HydroServer Lite becomes an endpoint for these services. Additionally, if a HydroServer Lite is registered in the HIS Catalogue, the data becomes discoverable and downloadable by applications such as HydroExcel and HydroDesktop (Ames et al., 2012). Additionally, because it follows WaterML encoding, open hardware data stored in the HIS can be included in international catalogues such as the Global Earth Observation System of System (GEOSS) catalogue. 3 RESULTS AND DISCUSSION A sensor was built and deployed in two instances to collect data every 15 minutes and transmit that data every hour. Data were collected for two variables: temperature in degrees Celsius and relative humidity. In the first deployment, 424 data points were collected and transmitted to the webserver over the ten days period. With sufficient sunlight in these days, the sensor was able to collect and transmit the data without losing power. In this deployment, the data, including the timestamp, were stored in a text file on the webserver. A graphical representation of the temperature and humidity data is shown in Figure 3. Figure 4 shows the maximum daily temperature recorded in the ten days of deployment Relative Humidity (%) Temperature ( C) 1/14 1/15 1/16 1/17 1/18 1/19 1/20 1/21 1/22 1/231/24/2014 Figure 3: Temperature and Humidity Values Collected by Sensor in First Deployment

6 Figure 4: Maximum Daily Temperatures (in degrees Celsius) Recorded in First Deployment In the ten day period, there were 29 instances in which the values for the previous hour were not sent. This represents 13% of the total transmission attempts. Of the 29 instances, a failure in consecutive hours only occurred twice. As an attempt to address this issue, a redundancy was introduced into the Arduino. With the redundancy, at each hour the GPRS modem transmitted the data for the past two hours. In this way, each set of data is transmitted twice in case there is a problem with one of the connections. In the second deployment, the transmitted data points were submitted to the server s PHP page which inserted them into the ODM and were thus viewable in HydroServer Lite. The data were plotted by HydroServer Lite and are shown in Figure 5. In the second deployment, due to lack of sunlight and time constraints, the sensor only collected data for a period of nine hours and in those nine hours, two consecutive hours worth of data were not received by the webserver. While the second deployment proved the automatic ingestion into HydroServer Lite was successful, further testing is needed to determine if the added redundancy will significantly improve transmission success. Figure 5: Temperature and Humidity Values Collected by Sensor in Second Deployment

7 4 CONCLUSION This paper presents an open hardware sensor and the automatic ingestion of its data into the CUAHSI HIS. This system is a low cost solution for gathering environmental data, viewing that data in near-realtime, and storing the data in a standards-based system allowing for easy access and sharing. Though in this paper only air temperature and humidity were measured and recorded, it is assumed that the system can be used to measure a number of environmental variables at low cost. 5 REFERENCES Ames, Daniel P., Horsburgh, Jeffery S., Cao, Yang, Kadlec, Jiří, Whiteaker, Timothy, & Valentine, David. (2012). HydroDesktop: Web services-based software for hydrologic data discovery, download, visualization, and analysis. Environmental Modelling & Software, 37, Anzlone, Gerald C., Glover, Alexandra G., & Pearce, Joshua M. (2013). Open-Source Colorimeter. Sensors, 13(4), Barniuk, Richard G. (2011). More is less: signal processing and the data deluge. Science(Washington), 331(6018), Borgman, Christine L. (2012). The conundrum of sharing research data. Journal of the American Society for Information Science and Technology, 63(6), Clasen, Thomas, Fabini, D., Boisson, S., Taneja, J., Song, J., Aichinger, E.,... Nelson, Kara L. (2012). Making Sanitation Count: Developing and Testing a Device for Assessing Latrine Use in Low- Income Settings. Environmental Science & Technology, 46(6), Conner, Lafe G., Ames, Daniel P., & Gill, Richard A. (2013). HydroServer Lite as and open source solution for archiving and sharing environmental data for independent university labs. Ecological Informatics, 18(0), doi: Fedi, A., Ferrari, D., Lima, M., Pintus, F., & Versace, C. (2013). Acronet Paradigm: An Open Hardware Project. Paper presented at the 2nd Open Water Symposium, Brussels. Hicks, S. D., Aufenkampe, A. K., & Montgomery, D. S. (2011). Sensor Networks, Dataloggers, and Other Handy Gadgets Using Open-Source Electronics for the Christina River Basin CZO. AGU Fall Meeting Abstracts, Horsburgh, Jeffery S., Tarboton, David G., Piasecki, Michael, Maidment, David R., Zaslavsky, IIya, Valentine, David, & Whitenack, Thomas. (2009). An integrated system for publishing environmental observations data. Environmental Modelling & Software, 24(8), Kadlec, Jiří, & Ames, Daniel P. (2012). Development of a Lightweight Hydroserver and Hydrologic Data Hosting Website. Paper presented at the Proceedings of the AWRA Spring Specialty Conference on GIS and Water Resources, New Orleans. Pearce, Joshua M. (2012). Building Research Equipment with Free, Open-Source Hardware. Science, 337(6100), Piwowar, Heather A., Day, Roger S., & Fridsma, Douglas B. (2007). Sharing detailed research data is associated with increased citation rate. PloS one, 2(3), e308. Tenopir, Carol, Allard, Suzie, Douglass, Kimberly L., Aydinoglu, Arsev Umur, Wu, Lei, Read, Eleanor,... Frame, Mike. (2011). Data sharing by scientists: practices and perceptions. PloS one, 6(6). Trevathan, Jarrod, Johnstone, Ron, Chiffings, Tony, Atkinson, Ian, Bergmann, Neil, Read, Wayne,... Stevens, Tom. (2012). SEMAT The Next Generation of Inexpensive Marine Environmental Monitoring and Measurement Systems. Sensors, 12(7),