We're sorry but this page doesn't work properly without JavaScript enabled. Please enable it to continue.
Feedback

Usability Engineering For Successful Open Citizen Science

00:00

Formal Metadata

Title
Usability Engineering For Successful Open Citizen Science
Title of Series
Number of Parts
183
Author
License
CC Attribution - NonCommercial - ShareAlike 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this
Identifiers
Publisher
Release Date
Language
Producer
Production Year2015
Production PlaceSeoul, South Korea

Content Metadata

Subject Area
Genre
Abstract
Citizen science can be explained as the engagement by nonprofessional scientists in collecting data, analyzing data, developing technologies and the publication of these on a voluntary basis. In a majority of citizen science projects the data to be collected is geospatial and is being presented on maps. If the data is about environmental observations, this approach is often referred to as participatory sensing. A novel approach in this field is to equip citizens with DIY-environmental sensor stations and to establish citizen driven sensor networks. This approach leads to better data coverage but also contains motivational aspects, as citizens build up their own devices. Most existing DIY sensor stations are not fully open in terms of source code, data collection, hardware, educational documents or extensibility for other platforms. That's why they lack transparency for the user, e.g. for scientists, who are interested in the data or citizens who want to understand the algorithms. The SenseBox project started at the Institute for Geoinformatics, University of M?nster, and is an ongoing open citizen science project. Based on open hardware components (Arduino microcontrollers and compatible sensors) citizens build their own Internet of Things enabled sensor stations to collect environmental data (temperature, humidity, air pressure, loudness, VIS-light, UV-light). The data is being published as open data and visualized on a web based platform, the OpenSenseMap (OSeM). An educational edition of the SenseBox and didactical material are being introduced into high schools, where students learn to code, measure environmental phenomena and work scientifically. The whole source code is open source, instructions are being published as open educational resources (OER) and models for a 3D-printed waterproof case are available as open source as well. In a first project phase, around 50 SenseBox stations were deployed to citizens and schools in Germany. Some participants had problems in the building and registration process of the SenseBox, others disconnected their SenseBox after some time. In this paper we want to investigate, if usability problems were the cause of the low success rate and how the motivation of citizen scientist can be preserved to ensure long-term data collection. In a user study we plan to evaluate the workflow of (a) building up the SenseBox station, (b) register it on OSeM and (c) verify the communication. Therefore we invite non professionals to our institute for an observed usability study in a controlled environment. A number of ten attendees is sufficient to improve usability significantly. As a basis, the current state of our online tutorials, containing text documents with illustrations and examples, are provided to the participants. Sessions are video recorded, and participants are being interviewed afterwards. In a next step, the results are analyzed to enhance project documentation and hardware design to a more user friendly version. The enhanced construction kits are being rolled out and provided to a new group of citizen scientists. In a final step, a questionnaire is used as an evaluation for the user study and the motivation of the participants.
126
WordStandard deviationGradient descentComputer hardwareCuboidShared memoryVarianceSoftware developerMeasurementRepresentational state transferWeb 2.0Object (grammar)TelecommunicationRun-time systemCombinational logicSocial classOpen setScaling (geometry)Group actionTexture mappingArithmetic meanOpen sourcePhysical systemMicrocontrollerSinc functionSource codeSoftware development kitBasis <Mathematik>PrototypeMereologyPressureThresholding (image processing)Limit (category theory)Computer programmingFood energyCartesian coordinate systemInformationWave packetBitServer (computing)Revision controlComputer animation
Computer hardwareUniverse (mathematics)Process modelingObservational errorSoftware testingComputer programmingNoise (electronics)Constructor (object-oriented programming)BitSoftware development kitOpen setUniqueness quantificationElectric generatorMereologyRepresentational state transferTexture mappingConnected spaceProduct (business)SoftwareCASE <Informatik>Touch typingResultantMultiplication signRun-time systemError messageImage registrationOpen sourceEmailDatabaseDiameterObservational studyMeasurementCuboidUsabilityMappingInsertion lossFlow separationRadiusSpecial unitary groupCellular automatonGreatest elementAxiom of choiceWordArithmetic meanProcess (computing)Expert systemLevel (video gaming)MicrocontrollerPhysical lawTerm (mathematics)Computer animation
Table (information)CuboidHome pageHypothesisStatisticsSoftwareConstructor (object-oriented programming)CASE <Informatik>Acoustic shadowUniform resource locatorExpert systemBuildingTelecommunicationWeb 2.0File formatMultiplication signMereologyDifferent (Kate Ryan album)Cartesian coordinate systemSoftware development kitComputer programmingRow (database)Product (business)Software testingPhase transitionArithmetic meanPresentation of a groupShared memorySoftware developerMeasurementGroup actionSign (mathematics)Process (computing)Price indexDependent and independent variablesLevel (video gaming)Moment (mathematics)Ideal (ethics)WebsiteArithmetic progressionPower (physics)Data storage deviceGame controllerLine (geometry)TwitterWeb pageDatabaseAsynchronous Transfer ModeComputer animation
Computer animation
Transcript: English(auto-generated)
Okay, thank you very much for the introduction and for the words. My name is Jan Wirvan. I'm from Germany from the Institute for Geoinformatics in Münster. My work is mainly about citizen science, which can be explained as the involvement
of the general public or non-professionals in scientific projects or in scientific development. In our case and in our development, we use crowdsourcing or the citizens collect information
for us, which are represented on maps, and this approach is called participatory sensing. So, a bit about my work at the Institute for Geoinformatics.
I'm working for the GI at School Lab, which is a school laboratory where we perform workshops for pupils and training for teachers mainly to bring GI science into schools and to teach
the teachers how to use GI tools like GIS, for example, in their school classes. On the other hand, the pupils learn how to work scientifically and to program, for example, to code or to build up sensor stations, which can be used then to gather
information which are then collected by our servers. This is also my part of the work there. I'm a pupil teacher who is working with open source hardware, so to say our basis
are microcontrollers, which you might know. It's an open source hardware platform, not only consisting of the microcontrollers, but also on a huge and still ungrowing community. There are a lot of applications for Arduino that are already there, and so you can find
a lot of inspiration there. Our work with the Arduino platform and the microcontroller programming has shown that it is really a playful and simple way to teach the basics of programming to the kids.
On the left-hand side, you see a very simple example, just as an example for an experiment where you can make an LED blink which is connected to a digital port of the microcontroller there.
Then, our idea was to collect the best experiments and projects into one toolkit. We provide the hardware and open source educational resources to schools to give schools
and teachers a toolkit for interdisciplinary work with DIY hardware and prototyping in their school classes. We are focusing on environmental sensing, so we are using temperature, humidity, pressure
sensors, but also air quality sensors, for example, that can be used to build up projects in the schools and in project days on schools, and then can be web-enabled to publish the
data and the measured data. After we collected all the parts for this sensebox toolkit, we thought about how to publish the measurements, and then we came up with the idea of the Open Sense Map,
which is a tool for sharing and publishing data. It started very simple, as you see here, the first prototype in 2013. We developed at our institute where it was only capable of displaying the real-time data,
so the last measurement of each of the senseboxes. The sensebox portal could be used in combination with the open API to connect the small sensor
stations that are built in the school classes to publish the measurements there. We came up with a new definition of project goals after we recognized that the demand
for senseboxes was very high, and we got a lot of requests from schools, and they wanted to use our system, so we thought to enhance our user group, that we can increase
participation not only by getting the data from the school classes, but also from other citizen scientists, other interested people who want to know something about their environment. For example, for a citizen, it could be nice to see how loud it is in the street, or how
the air pollution behaves, for example. We want to establish a large-scale sensor network, and the new version of our Open
Sense Map is that we want to be a new standard for open sensor data for sharing and collecting open sensor data in Germany. We used a bit of an enhanced concept of the sensebox here for the hardware infrastructure,
which is now a web of things enabled sensor station, so that means that it is a physical object integrated into the web, and this can be addressed and identified by a specific URI, and the communication of the sensebox to the sensebox or between the senseboxes
can be achieved by an RESTful API, so to say, by HTTP methods for getting measurements, for example, or posting new measurements to the server. You can see in the picture that there is not only the communication and sensing part,
but also the acting part, so that means that you can, for example, specify thresholds and get warnings if you reach, for example, a certain temperature limit, or if the noise
pollution is too high, you can get warnings from that. On the other hand, the Open Sense Map serves as a server infrastructure for our sensor network, and we want to keep all the data open, so all data that you upload there is
published as open data and can be shared between communities or scientists or interested people, and as I said, it is open in all stages, so our RESTful API is designed for
open access, which means that you don't need a sensebox for a contribute in our campaign, but you can also create your own sensor station, for example, with the Raspberry Pi or with another web-enabled sensor and implement the methods to connect to our API to get any
sensor working on the Open Sense Map. Another feature is that if you use the sensebox and you register it, you can have an automated sketch generation.
A sketch is a software program for those auto-ino microcontrollers, and this is especially useful if you are a non-expert and don't want to implement all the methods to connect to the API by hand. So, what we did was we did a test run in 2014 in April, and then we started to equip
citizens with our do-it-yourself construction kits, which were based on those senseboxes
for schools, but made a bit more simple to really get it running for non-experts as well.
And the stations, we tried to distribute them mainly all over Germany. Here you see the biggest cluster, which is around Munster, where our university is located. Of course, we have most of the test boxes placed up there.
And I forgot to say that we are trying to use low-cost hardware parts for it, because if a station would cost several hundred dollars or maybe several thousand dollars, it would be a big disadvantage for our campaign, because we want to achieve a coverage for
really all over Germany, and therefore we need very high participation, and this could only be achieved by low-cost products. So, of course, we cannot compete with high-quality sensor stations from the government, for example,
which costs, when a quality station, for example, in Munster, costs over 30,000 euros, and this is, of course, not affordable for a citizen or for a volunteer.
But there's only one station for measuring air pollution in Germany, about 200 km diameter from the sound center.
We can say, okay, we can't put a very accurate and very expensive measuring station there, but we can put several hundreds of stations in the same radius, and so can reduce the measurement error by every new station which is registered in our portal.
Yeah, and the first results, after one year, after around one year, we had around 8 million measurements in our database at the Open Sense Map server, but a lot, and I must say, most
of the stations were disconnected after a few weeks, so we had to write emails to each of the users and say, hey, can you reset your station, or what is going on, but this was most of the times not answered by them, so yeah, we thought about why did that happen and how we can find the error sources.
So what we did was we performed a user study in a controlled environment at our institute and we invited volunteers to take part in a user study and we tested the hardware wiring,
the software installation, the registration process, and the documentation material and checked if everything is understandable also for a non-professional user who may have never gotten in touch with open source hardware or sensor equipment or anything like
that, and the result was, after a couple of weeks of user studies, the new SenseBox, a more user-friendly and more simple SenseBox, and you can see it here on the right side,
it's really looking like a product now, and this is now an environmental DIY sensor station consisting of a sensor, of a microcontroller, of a network adapter, of specially designed
hardware part for simple hardware connections where you don't need to wire everything up, but you can just use cable with unique connectors to connect the sensors to the microcontroller, and we tried to cover the values of interests.
We found out in a questionnaire of the citizens which they are interested most in, and these are temperature, humidity, air pressure, noise level, and illuminance, and the UV
level, which is especially useful for reducing skin cancer risk if you know about the UV radiation intensity. The station has to be built up, deployed once, and is then continuously measuring
data, and all the data, as I said before, is published as open data. Here you see the inside of our construction kit.
I also have one with me if you are interested, I can show it to you later, and we also provide a 3D model for a case for the sensors, consisting of lamella for protection against
radiation of the sensors, and a waterproof box where the main station is placed. What are we doing next? We want to do a second test run where we deploy around 100 further stations and see
if we can now achieve the continuous measurements without the need to restart the stations, for example, or without any problems. Then we start our official measurement campaign where we sell the sensor do-it-yourself boxes
in an online store, and then extend our documentation continuously, and we also want to include the public, the creative community in the development so that, for example, citizens
find new use cases or want to upgrade other sensors, they can make manuals, put them online on our website, on our web portal, and share it with others. What is another very important part is to find partners and promote the project.
We already did that by visiting fairs and exhibitions, for example, or inviting journalists or institutes to make our project more known in the outside world.
If you have any questions or ideas, I am happy to speak about it, and thank you for your attention. Thanks, John. Thanks for an interesting presentation.
Do you have any questions from the audience? At the moment, we are working in Germany, but the main reason because of that is because we are founded by the Federal Ministry of Education and Research in Germany, and they
are mainly interested in the sensor network for Germany, but that doesn't mean that we aren't open for other projects, but as we are only a very small team and need the support of the ministry, we mainly have to take care about Germany at the moment,
I would say. I think in about maybe half a year when we stopped our second test phase and we have a ready product which is working in all stages, then we can provide the sense box also to other institutes or interested people, mainly everywhere.
I was asked to use a microphone for questions and answers for the recording purposes. The original sense box, as far as I can see, has some educational values in actually
building the devices by kids, that the kids would learn about programming and building devices in the process, so in the new design, do you think that this aspect of educational value is diminished, and did you worry about this aspect or take extra care to prevent
this? Yes, this is a good question. We have, besides the construction kit for the citizen science purpose, we have another construction kit for the schools, so the construction kit for the citizen science application
is called sense box home, and the other for the school application is sense box edu for the educational part, and the difference between both of the boxes are that the sense box for the citizen is much more easy to build up, and the sense box for the schools,
the sense box edu is much more going into the basics of programming, of electronics, and so, as I said before, we tried to enhance our use case in the sense that we want to
have a larger user group, so we developed the citizen science part after the educational part. We have a couple of minutes for other questions. We have time for a couple more questions, and I'll ask one or two.
So your data is open, and when you transfer data, what format do you use for the data?
We are using for the communication geojson format to get it into the database, and you can not only get the data by our API, as I said before, but you can also download them as CSV, for example, or also as JSON data, so that also somebody who doesn't want
to implement anything can make use of the data and get a table from our home page from the station he selected and can download the data from there. I think this is a very good example of crowdsourcing tool, and one of the drawbacks
of crowdsourcing tool is validation, how to validate your data. Yeah, this is right, and this is a really big problem. What we are thinking about is first, after you build up the station, it has to be calibrated,
so what we want to do is to make a calibrating session where people of a certain location can visit, for example, the town center where one expert can take care of the calibration,
but what is a big, or what can be a disadvantage is that everybody who is putting the sense box and set it up somewhere is maybe placing this into the sun or into the shadow so we
get different measurements, but we have several, for example, bachelor thesis about that where we use statistical methods to filter out stations which deliver wrong data or which are not configured properly or software that can be used to determine if the station
is set up inside or outside, something like this is in progress, but this is a problem and we have to further develop there.
Thank you very much, Jan. Thank you.