IN
DEGREE PROJECT CIVIL ENGINEERING AND URBAN MANAGEMENT,
SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2019,
Understanding and Improving the Utilization of Web and Mobile GIS Solutions for Outdoor
Environment Management
LEA ZUNA
Abstract
Trees, plants, water, playgrounds, green areas and benches are just a few out of many components that forms a pleasant outdoor environment. Not only are management and maintenance necessary to keep the environment clean and nice for everybodys well-being, but also documentation and records of geographic data representing the outdoor environment. Management and maintenance requires plans and cost calcula- tion where the geographic data is used as a basis. The data needs to therefor be reliably and continuously controlled and updated, especially due to the components of an outdoor environment constant change and move. An effective way to store, visualize and manage data related to geographic locations is to use Geo- graphic Information System (GIS) solutions.
This thesis is applying a user-centered design approach to derive a prototype that encourages use of web and mobile GIS solutions in the field during controls and maintenance and thereby resulting in more efficient data update procedures and maintenance operations. A case study is performed involving the two housing companies Svenska Bost¨ader and Stockholmshem where management and maintenance of geographic data representing the outdoor housing environment is involved in their work and is today managed by primitive methods. The user-centered design approach Focus Rapid Contextual Design is applied to derive user re- quirements in terms of functionality and user interface for further use to develop and integrate a prototype within the housing companies existing GIS solutions. The overall objective is to investigate, document and evaluate how an user-centered design approach can be applied to introduce web GIS solutions that will improve the workflow between office and field, this by reducing the number of steps performed during a field control by replacing primitive tools such as paper and PDF maps used today with the suggested GIS solutions. The potential long term impact of this study is to encourage the layman users that are working in field to use these new solutions instead of primitive methods, this by adapting functionalities and user interface in the web and mobile GIS solutions according to the layman users needs and levels by applying the user-centered design approach.
The Focus Rapid Contextual Design method shows to be successful in terms of gathering and manag- ing user data in a structured way. Due to tight time schedule and project resource limitations, an iterative prototype process where user feedback is used to redesign the prototype for further improvement was not performed. Only one prototype evaluation was performed and did not take place in context of use. This lead to weak user feedback meaning that the prototype most likely could have been further adapted to the users needs if an iterative process that took place in context of use was performed. It would most likely facilitate the faster adaptation for the web GIS solutions in the housing companies daily work. Encouraging layman users to change from manual to digital systems is seen as one if the major challenges during this thesis. Even though a user centered design approach contributes to reach the potential long term impact of this study which is to encourage layman user to use the suggested solutions, most likely organizational changes and policies are needed to accomplish process changes and make field workers change their work- ing manners. A small case study like the one performed in this thesis could possibly have been performed excluding the Working Models and Storyboarding in the Focus Rapid Contextual Design method with- out affecting the final result, this to gain more time for the iterative prototype process as well as time for involving more users. Doing so could improve the work in this study to better reach the stated goals.
Sammanfattning
Tr¨ad, v¨axter, vattenomr˚aden, lekplatser, gr¨onomr˚aden och b¨ankar ¨ar bara ett f˚atal av m˚anga andra kompo- nenter som formar en utomhusmilj¨o. Inte ¨ar bara sk¨otsel och underh˚all n¨odv¨andigt f¨or att h˚alla utomhus- milj¨on ren och trevlig f¨or allas v¨albefinnande, utan ¨aven dokumentation och protokoll ¨over den geografiska data som representerar utomhusmilj¨on. Sk¨otsel och underh˚all kr¨aver planering och kostnadsber¨akningar d¨ar geografiska data st˚ar till grund. Data m˚aste d¨arf¨or vara p˚alitlig, kontinuerligt kontrollerad och uppdat- erad, speciellt i avseende till att komponenterna i utomhusmilj¨oer ¨ar i en konstant f¨or¨andring och r¨orelse.
Ett effektivt s¨att att lagra geografiska data p˚a ¨ar att anv¨anda sig Geografiska Informationssystems (GIS).
I denna studie appliceras en anv¨andarcentrerad designmetod f¨or att erh˚alla en prototyp som uppmuntrar anv¨andning av webb- och mobil-GIS ute i f¨alt i samband med kontroller och sk¨otsel och d¨armed resulterar i mer effektiva datauppdateringsprocedurer och sk¨otseloperationer. En fallstudie ¨ar utf¨ord och involverar bostadsf¨orvaltningarna Svenska Bost¨ader och Stockholmshem d¨ar sk¨otsel och underh˚all av geografiska data som representerar bostadsf¨orvaltningarnas utomhusmilj¨o ing˚ar i deras arbete och sk¨ots idag med hj¨alp utav primitiva metoder. Den anv¨andarcentrerad designmetoden Focus Rapid Contextual Design anv¨ands f¨or att erh˚alla anv¨andarbehov g¨allande funktionalitet och anv¨andargr¨anssnitt f¨or att vidare kunna utveckla och integrera en prototyp i bostadsf¨orvaltningarnas redan existerande GIS-l¨osningar. Det ¨overgripande m˚alet ¨ar att unders¨oka, dokumentera och utv¨ardera hur en anv¨andarcentrerad designmetod kan appliceras f¨or att introducera web-GIS l¨osningar f¨or att f¨orb¨attra arbetsfl¨odet mellan kontor och f¨alt, detta genom att reducera antal steg som utf¨ors under en f¨altkontroll genom att ers¨atta primitiva verktyg s˚a som pappers- och PDF-kartor som anv¨ands idag med f¨orslagna GIS l¨osningar. Den potentiellt l˚angsikta effekt av denna studie ¨ar att uppmuntra oerfarna anv¨andare som arbetar ute i f¨alt att anv¨anda dessa nya l¨osningar ist¨allet f¨or de primitiva l¨osningar som anv¨ands idag, detta genom att anpassa funktionaliteter och anv¨andargr¨anssnitt efter anv¨andarnas behov och niv˚a genom att till¨ampa den anv¨andarcentrereade designmetoden.
Focus Rapid Contextual Design metoden visar sig vara effektiv i samband med insamling och hanter- ing av anv¨andardata och utf¨ors p˚a ett strukturerat vis. P˚a grund av ett tight tidsschema och en resurs- begr¨ansning under denna studie ¨ar den iterativa prototypprocess som f¨oresl˚as i metoden inte applicerad.
Prototypen utv¨arderades enbart en g˚ang av anv¨andarna och utf¨ordes inte ute i f¨alt. Detta ledde till bristf¨allig anv¨andarfeedback vilket inneb¨ar att den slutliga prototypen m¨ojligen skulle kunnat anpassats ytterligare efter anv¨andarnas behov om man hade utf¨ort den iterativa prototypprocessen. Detta skulle troligtvis ocks˚a bidra till en snabbare anpassning av webb-GIS l¨osningarna i deras dagliga arbete. Att uppmuntra oerfarna anv¨andare att byta fr˚an ett manuellt till ett digitalt system ans˚ags som en av de st¨orre utmaningarna un- der denna studie. ¨Aven om den anv¨andarcentrerade designmetoden bidrar till att uppmuntra anv¨andare att anv¨anda de nya l¨osningarna, beh¨ovs troligtvis organisationsf¨or¨andringar och policys f¨or att uppn˚a pro- cessf¨or¨andringar och f˚a f¨altarbetare att ¨andra sina arbetsvanor. En s˚a pass liten studie som utf¨ors h¨ar skulle m¨ojligen kunna utf¨oras utan processerna Working Models och Storyboarding i Focus Rapid Contextual De- sign metoden utan att p˚averka det slutliga resultatet avsev¨art. Detta skulle bidra till mer tid f¨or att till¨ampa den iterativa prototypprocessen och involvera fler anv¨andare vilket troligtvis skulle f¨orb¨attra arbetet i denna studie f¨or att b¨attre kunna uppn˚a de satta m˚alen.
Acknowledgments
Emil Plynning, Digpro AB, co-supervisor. For always being there when needed and providing help and ideas as well as great support to reach deadlines in time.
Associate Professor Gy˝oz˝o Gid´ofalvi, KTH Geoinformatics, supervisor. For providing great feedback and ideas throughout the work of this thesis.
Professor Yifang Ban, KTH Geoinformatics, examiner. For critique and examination of this thesis.
Sofie Rahm, Digpro AB, senior supervisor and Daniel Sedell, Digpro AB. For feedback, inputs and shar- ing extensive knowledge.
Johan Winther, Digpro AB. For providing help with the start-up phase of this thesis.
Finally thanks to Anna Svahlstedt (Svenska Bost¨ader), Fatima Durrani (Stockholmshem) and Lars Berg (Stockholmshem) for participating in interviews and sharing important knowledge and experiences regard- ing the daily routines of the outdoor environment controls.
Contents
Abstract 2
Sammanfattning 3
Acknowledgements 4
1 Introduction 7
1.1 Background . . . . 7
1.2 Management of Outdoor Environment . . . . 8
1.3 Objectives . . . . 8
1.4 Scope and Limitations . . . . 9
1.5 Previous Studies . . . . 9
1.6 Disposition . . . . 10
2 Related Theories and Technologies 10 2.1 Web and Mobile GIS . . . . 10
2.2 Field Based GIS . . . . 11
2.3 User Experience and User Interface . . . . 11
2.4 Technical Background of Used GIS solutions . . . . 13
2.5 Used GIS Solutions and Modules . . . . 14
3 Methodology 16 3.1 Choice of Method . . . . 16
3.2 Focused Rapid Contextual Design . . . . 18
3.2.1 Contextual Inquiry . . . . 18
3.2.2 Interpretation Session . . . . 19
3.2.3 Affinity Diagram and Data Consolidation . . . . 20
3.2.4 The Wall Walk and Visioning . . . . 20
3.2.5 Storyboarding . . . . 21
3.3 Prototyping . . . . 21
3.3.1 Configuration . . . . 22
3.3.2 Evaluation . . . . 22
4 Results 22 4.1 User Requirements . . . . 23
4.1.1 Affinity Diagram . . . . 23
4.1.2 Visioning . . . . 25
4.2 Prototype . . . . 25
4.2.1 User Accounts and Roles . . . . 25
4.2.2 Informative and Easily Read Maps . . . . 27
4.2.3 Accurate and Effective Work Flow for Map Documentation . . . . 29
4.3 Workflow . . . . 32
4.4 Evaluation . . . . 33
5 Discussion 34 5.1 Implementation of Contextual Design . . . . 34
5.2 Prototype Evaluation . . . . 34
5.3 The Resulting Web GIS Prototype . . . . 35
6 Conclusions and Future Work 35
6.1 Conclusions . . . . 35
6.2 Future Work . . . . 36
6.2.1 Suggested Prototype Improvements . . . . 36
6.2.2 Further Development of the Suggested Web GIS Solution . . . . 38
References 39
1 Introduction
Trees, plants, water, playgrounds, green areas and benches are just a few out of many components that forms a pleasant outdoor environment. Not only are management and maintenance necessary to keep the environment clean and nice for everybodys well-being, but also documentation and records of geographic data representing the outdoor environment. The geographic data is used as a major basis for cost cal- culations and outdoor environment planning. An outdoor environment is constantly changing and it is therefore of high importance to perform controls to make sure that the data is updated. Geographic data can be managed by a Geographic Information System (GIS) which is a system used for storage, update, visualization and more of geographic data. This thesis aims to investigate how to improve existing and underutilized web and mobile GIS solutions for possibilities to be used in terms of data collection and up- date in field during outdoor environment controls and maintenance. This study contributes with proposed GIS functionalities to encourage layman users to use GIS solutions for data collection and update in field instead of primitive methods such as static paper or PDF maps. Factors that discourages the user to use the existing GIS solution and user requirements are identified by using a user-centered process to gather user data via field studies, rationalizing workflow and designing user interfaces. This is realized by a proof of concept prototype to demonstrate the main functionality of the idea. This section gives an introduction of this study by first introducing its background and rational in Section 1.1 followed by outdoor environment management in Section 1.2. Research objectives are stated in Section 1.3, scope and limitation in Section 1.4 and previous studies in Section 1.5. Finally a disposition is presented in Section 1.6.
1.1 Background
Svenska Bost¨ader and Stockholmshem are two housing companies owned by Stockholms stad, where man- agement and maintenance of outdoor environments are one of their major tasks, involving many geographic locations as well as participants. Both companies are using Geographic Information System (GIS) solu- tions, provided by the same GIS provider, for management of outdoor environmental data pertaining to geographical objects or phenomena such as grass, roads, trees, snow, playgrounds and more. Unfortu- nately, only a few out of many possible employees are using the GIS solutions, and GIS in general, in their daily work. A brief look into the companies working processes shows that today the GIS solutions are heavily under-utilized and primarily used as a view port rather than a tool for decision making. This under-utilization is also present in fieldwork conducted, as instead of using a web and mobile GIS solution, the companies use printed paper maps along with inventory lists and paper notes. The fieldwork is upon completion digitized, but the work has to be more or less done twice as everything has to be manually entered into the GIS solution. Not only does this lead to inefficiency, but also to potential risks for addi- tional errors sources, such as ambiguities in writing and drawing interpretations. Efficiency and additional data management error sources in fieldwork are brought up in Design, Implementation and Evaluation of a Mobile GIS Solution for a Land Registration Project in Lesothoby A. Bronder and E. Persson [1].
Two major problems that the companies seems to struggle with are 1) lack of knowledge within the field of GIS and 2) out of date GIS solutions. In the article GIS - ett outnyttjat verktyg inom kommunerna (GIS - an underutilized tool within municipalities), by S. Edsg˚ard [2], it is stated that the Swedish municipalities are good at collecting data, but are not as good at using it for analysis or decision making, where even here the lack of knowledge is seen as one of the major reasons. Concerning Svenska Bost¨ader and Stockholmshem for instance, they can be considered good at collecting data in the sense that the data is collected and up- dated regularly, although they are not very efficient in doing so. This data is sparsely used after collection, stored in different formats and as stated earlier, is mostly used for viewing purposes. Possible GIS solutions out in the field, to help with data collection and decrease manual labor, is not being considered at all today.
This is most likely due to lack of knowledge about possible solutions, giving the impression that the cost of the known solutions to the organization are perceived to outweigh the benefits of implementing them.
1.2 Management of Outdoor Environment
An outdoor environment is formed by many components where many of them are in a daily change. It is of high importance to maintain and keep the environment clean and nice for everybody well-being, as well as keep records and document it. Svenska Bost¨ader and Stockholmshem are managing and documenting their outdoor environment by using GIS solutions. Both companies are using dpMap and dpWebmap, two GIS solutions provided by Digpro AB who develops innovations in Geographic Information Technology.
dpMap is a complex desktop application used just by a few employees at the housing companies to store, update and analyse data, while dpWebmap is web and mobile application available for more employees and is used as a view and information tool. Both applications are equipped with the module Area Maintenance which is specifically focusing on the geographic outdoor environment data and makes it possible to map spatial data representing trees, green areas, roads and more.
The maps are used as a base to create management plans, cost calculations and during the outdoor en- vironment maintenance performed by hired maintenance entrepreneurs. These maps needs to be regularly controlled and updated since the outdoor environment is changing constantly. Svenska Bost¨ader did for instance in 2015 renovate a whole outdoor environment area where they planted 56 new trees, 674 roses, 5449 bushes, 13 715 perennials and 19 770 spring onions. All these needs to be included in their cost and management calculations as well documented in their outdoor environment map system [3].
The field controls and data updates are today managed manually by either using paper maps to make notes on of observed changes while out in the field, or by using dpWebmap as a viewing tool and make notes on PDF maps of observed changes. The notes are further transcribed to make it possible for others to interpret during the data insertion. The notes are further sent to a main responsible to perform the map data update by using dpMap. Non of the companies are using field-based GIS solutions to update the digital map while in the field. Doing so instead could save time, money and lead to a more accurate and efficient data update.
1.3 Objectives
The aim of this study is to apply an user-centered design approach to derive a prototype that encourages use of web and mobile GIS solutions in the field and thereby resulting in more efficient data update procedures and maintenance operations.
The overall objective is to investigate, document and evaluate how an user centered design approach can be applied to introduce web GIS solutions that will improve the workflow between office and field, this by reducing the number of steps performed during a field control by replacing primitive tools such as paper and PDF maps used today with the suggested GIS solutions.
The potential long term impact of this study is to encourage the layman users that are working in field to use these new suggested solutions instead of primitive methods, this by adapting functionalities and user interface in the web and mobile GIS solution according to the layman users needs and levels by applying the user-centered design approach.
To accomplish the stated goals, a case study is performed involving the two housing companies Sven- ska Bost¨ader and Stockholmshem where management and maintenance of geographic data representing the outdoor housing environment involves their work. The Contextual Design method is applied to gather user requirements which are used as a base during the prototyping that is implemented in web and mobile GIS solutions owned by the housing companies today. The final result is evaluated by end users through questioners where the feedback is used for further improvements by adapting the user centered design approach.
1.4 Scope and Limitations
The Contextual Design is a method involving many participants, both in terms of field interviews and data evaluations. Each step should be performed including at least two people, in some steps even larger groups with different expertise areas related to the product evaluated. Some steps are suggested to perform within a few hours, some over a time span of couple of days. Due to limitations in both the time frame and number of people available, the steps of the Contextual Design is scaled down in terms of both. Since the prototype is based on existing web GIS solutions and functionalities, there are some restrictions in terms of functionality development and improvements of the UI design, which are rather discussed than included in the prototype. The layman user is in focus, and therefor even simplicity and usability. More complex solutions and ideas are brought up in Section 6.2.2. This study is restricted to only look into the housing companies web GIS clients, even though the desktop GIS clients are a major part of their new suggested workflow.
1.5 Previous Studies
Several similar studies have been conducted with the aim to develop web-based systems by using different user-centered approaches to identify layman user requirements and realize these by proof of concept proto- types, where the prototypes are further evaluate by diverse approaches. In [4] by F. Hilding and E. Syk, a prototype of a web based Network Information System (NIS) is created, where the layman user is in center of the entire design. The prototyping is based on interviews with the users of the NIS system as well as studies of scientific and technical literature. The prototype is evaluated by a survey collecting quantitative data along with a demonstration session based on feedback from both experts and layman user in the field.
T. Fledderus is in [5] in similarity to this study implementing the Contextual Design method to create a design hypothesis with focus on learnability and effectiveness which is two out of five qualitative com- ponents in terms of usability according to J. Nielsen. In comparison to the study presented in this thesis, where the final result is only evaluated ones, the final result in T. Fledderus study is evaluated in three iter- ations where the resulting feedback is used to build the final design hypothesis. G. Ekstedt and T. Endoff aims in [6] to develop a mobile GIS tool with same purposes as in this study - to replace paper maps used during conduction of fieldwork and field inventory work in the Swedish municipalities. Human-Computer Interaction and usability goals are in focus during a litterateur study along with studies of already existing similar applications. C. Ekman and J. Martinsson are in [7] applying a modified version of the Contextual Design method named Focus Rapid Contextual Design, which is the same approach used in this study, with the aim to evaluate the method itself and to recommend alterations based on team properties, time frame and available resources with focus on small teams of two members with limited resources. A smartphone app development is realized by a digital prototype, where an evaluation is performed to see if the paper prototype suggested in the Contextual Design process could possibly be replaced or complemented by dig- ital prototypes, which was an approach used in the case of this study.
Using user-centered approaches in terms of developing web-based systems for layman users seems promis- ing in gathering user requirements as well to develop prototypes [4–7]. It seems though to be a chal- lenge in terms of evaluating the final results in projects with tight times schedule and engage relevant end-users [4, 5]. C. Ekman and J. Martinsson states that it is a long process to start using an application and make a user acquainted with it, which yields to difficulties to complete a full evaluation [6]. E. Syk and F. Hilding as well as T. Fledderus states in the final results that user-centered design process depends heavily on the developers understanding of the users [4, 5] and it is important to not fill the tool with unnecessary features, functions or choices [4]. The study of C. Ekman and J. Martinsson shows that the Focus Rapid Contextual Design method has a potential to work well with smartphone app development, especially when the objectives are vague and usability in focus. This study shows that two of the steps performed could be skipped in terms of smaller projects; the Data Consolidation and the Storyboarding, and that the paper prototype could be replaced by a digital prototype, even though both the digital and
paper prototypes has its pros and cons. Since the Rapid Contextual Design is defined in 2005, where it is suggested to use a paper prototype also in terms of time limitations, it is suggest by the authors to overlook the prototyping component and update it since the possibilities to create digital prototype has increased during the last ten years and may be quick to use in terms of small projects [7].
1.6 Disposition
Section 2 presents related theories and technologies to this study. The choice of methodology and its application is presented in Section 3 followed by the results in Section 4. A discussion is performed in Section 5 where final conclusions and recommendations for future work are drawn in Section 6.
2 Related Theories and Technologies
This section covers the concept of web and mobile GIS in Section 2.1, and describes further how the system can be used in field for data collection in Section 2.2. The definition of User Experience, UX, and User Interface, UI, if described in Section 2.3. Section 2.4 covers the technical background of the housing companies already owned GIS solutions, where these GIS solutions and additional modules are further presented in Section 2.5.
2.1 Web and Mobile GIS
Geographic Information System (GIS) is a computer-based information system with functionalists such as collection, storage, processing, analysis and visualization of geographic data. Software, hardware, data, people and analysis/methods are important components of geographic information handling. By combin- ing these components in human and internet networks, as seen in Figure 1, ideas and information can be spread through the networks for a better understanding and improvement of the world [8].
Figure 1: Important components in a GIS network. Source: [9]
According to Environmental System Research Institute, ESRI, an international supplier of GIS software, web GIS can be defined as any GIS that uses web technology to communicate between web clients and servers [10] with following key elements essential to web GIS:
• The server has a URL so that clients can find it on the web.
• The client relies on HTTP specifications to send requests to the server.
• The server performs the requested GIS operations and sends responses to the client via HTTP.
• The format of the response sent to the client can be in many formats, such as HTML, binary image, XML (Extensible Markup Language), or JSON (JavaScript Object Notation).
In comparison to traditional desktop GIS, which is usually used by trained expert users, web GIS can be used by hundreds of users simultaneously including many novice users. The web and mobile GIS requires therefore higher performance, scalability and usability than traditional desktop GIS [10]. The web and mobile GIS tools area often adapted to usage and users, so a tourist for instance will require quite different options in comparison to maintenance staff [11]. D. Browning states in [12] that the growing proliferation of mobile devices and the displaying of GIS data on the web have been changing the GIS industry. Collection of GIS data using mobile devices and display that data on the web is increasing rapidly within organizations due it ease of use and the ability for the end-user to easily collect and distribute data to the world. Not only is the web an interesting medium for geospatial data presentation due to its platform independency and possibility to reach many users at a minimal cost, but it also allows for dynamic and interactive maps with new mapping techniques that are not possible with traditional paper maps [13] .
2.2 Field Based GIS
The web and mobile GIS is frequently used for data collection in field, which makes it possible to transform the traditional paper workflow to efficient digital workflow. There are many field data collection application available on the market within several field areas. To mention a few, collector for ArcGIS is for instance a mobile data collection app which makes it easy to capture and edit accurate data and return it to the office. It integrates with ArcGIS and is possible to use offline [8]. Another field based GIS application is ”GEOSECMA Park och Gr¨onyta” provided by S-GROUP Solutions, which is a system used to manage and document the green infrastructure [14]. Mobile Data Collection is another field-based GIS application provided by GIS Cloud for real time data collection, visualisation and monitoring which is then possible to share with team members [15]. Depending on the purpose of work, there are plenty to chose among.
2.3 User Experience and User Interface
User Experience, UX, and User Interface, UI, are two frequently used terms in computer science and web applications and familiar to many. What differs one from the other though may be vague. The definition of UX according to The Interaction Design Foundation is ”The process of creating products that provide meaningful and relevant experiences to users. This involves the design of the entire process of acquiring and integrating the product, including aspects of branding, design, usability, and function”[16]. Accord- ing to the same foundation, the definition of User Interface is ”The process of making interfaces in software or computerized devices with a focus on looks or style. Designers aim to create designs users will find easy to use and pleasurable. UI design typically refers to graphical user interfaces but also includes others, such as voice-controlled ones”[17]. Differences between UI and UX are briefly described in Figure 2.
Figure 2: Difference between UI and UX design. Source: [18]
The UI is a subset of UX and according to P. Morville, a designer and information architect who has been working in this field since 1994, the UX is described with seven factors; Useful, Usable, Findable, Credible, Desirable, Accessibleand Valuable [19]. UX Design goes beyond UI Design and considers why, what and how of product use [16] while UI is seen as a craft involving the building and essential part of the user experience [17]. The UX design is a user-centered and iterative process, meaning that the design is based on the understanding of the users and their context. Using that understanding, requirements are stated upon which the suggested design solutions are built. An evaluation of the product is conducted against the requirements and an iterative process takes place. The feedback obtained from evaluation is used to improve the product, which then is followed by an additional evaluation. So it continues until all the relevant issues and user needs are addressed optimally [16]. The user-centered process is drawn in Figure 3.
Figure 3: Representation of a User-Centered process. Source: [19]
UX is of high importance during the development process interactive maps and web GIS solutions. Maps today are highly interactive and delivered online, usually through a mobile device and the boundaries be- tween a map maker and a map user is dissolving. In The Geographic Information Science and Technology Body of Knowledgeby John P. Wilson core concepts from UI/UX design that is important to cartography and visualization are introduced. Some UI/UX recommendations can be found in common operator prim- itives in cartography and visualization, such as overlays, symbolization, zooming, panning, re-projecting, searching and more [20].
2.4 Technical Background of Used GIS solutions
dpSpatial is the heart of the GIS system which is a solution that integrates advanced technology with Or- acle Spatial or PostgreSQL databases for storing, analysing and presenting geographical data. The whole system is web-based which leads to a reduction in expensive administration and client installations. As seen in Figure 4, a typical installation of dpSpatial consists of one database server and one or more Tomcat servers. A Tomcat Server is an open-source Java Servlet Container and provides a Java HTTP web server environment in which Java code can run. A Java Database Connectivity, JDBC, is used which is a Java- based data access technology used for Java database connectivity. Depending on the size of the installation, a failover database server and a load balancer may used in front of the Tomcat servers.
The dpSpatial base patform consist of three major client types:
• The Java client with full editing functionality and powerful analysis functions
• The dpWebmap client which can be used on any modern browser or mobile device
• External systems that use dpSpatial through integrations/APIs
The dpSpatial system consist of objects and components stored in the database. An object usually repre- sents something physically, such as a building, wheres components usually are representations of an object, such as a geographical object in the map (a polygon, line, or point) or a data component with attributes like area of the building, owner, building year and more. Each component is represented by a table in the database and each created component is stored as a row in that database table. Each object can be represented by many components and one component can belong to many objects, such as for instance a geographical line which can be a component of both a road and a pipe line. Subtypes can as well be used to differentiate between objects which have the same database structure but have different permission, drawing attributes and so on. In terms of a building, it could have the three subtypes apartment, greenhouse and villa. It also possible to connect code lists to attribute fields. A code list is a database table consisting of different attribute values. Instead of having the possibility to type in any owner in an attribute field, it is possible to choose among a few specified values in a list instead. A simplified graphical object model is drawn in Figure 5.
Figure 4: The system behind dpSpatial base platform. Source: Digpro
Map products in dpSpatial are further used to specify what objects to show, at what scale each object should be drawn and what drawing attributes that should represent each object in a map layer. A building could be specified to be represented by a yellow polygon in a scale 0:100000 whilst a tree could be represented by a tree symbol colored in green in a scale 0:10000. A map product is used as a base when creating a Web Map Service (WMS) or Web Feature Service (WFS).
For each object it is possible to specify what attribute fields that shall be visible in their respective at- tribute form and how it should be designed. It is also possible to define which attributes that should be editable and not.
When a user is working in dpSpatial, the work is always done in something called a “workset”. A workset is a working set, in which all changes made in the map are saved and only visible in that particular workset and for that particular user. When the workset is posted/committed, the changes are made in in the original map and are possible for other users to see. Each user can have as many worksets as possible, and it pos- sible to send worksets between users. For each component table in the database, two additional tables are used to store the contents of worksets and a history of previous changes.
Figure 5: A simplified graphical database object model used in dpSpatial.
2.5 Used GIS Solutions and Modules
The GIS solutions used by Svenska Bost¨ader and Stockholmshem for their outdoor environment manage- ment are dpMap and dpWebmap. Both applications are using the dpSpatial base platform and are running against the same Oracle database, in which all spatial and non-spatial outdoor environment data is stored.
dpMap is one of Digpros many Java client applications, mostly used by the municipalities and admin- istrations for municipal planning and management. Only the main responsible at the housing companies does have permission to use dpMap, which is the main tool used for storing, analysing and presenting geo- graphical data. This is a complex tool where normally training is required for possibilities to use it. dpMap
is downloaded and stored locally on the device and is not possible to run in a browser, it is therefor usually referred to as a ”desktop client”. Unlike dpMap, dpWebmap is a web and mobile client possible to run in any web browser with a responsive interface suitable for all platforms. It is an OpenLayer-based client and can be configured to present geospatial data from different geodata services such as WMS for raster data representations and WFS, or Keyhole Markup Language, KML, for vector representations. dpWebmap is a simple GIS tool with functionalities mostly used for visualization, but it does support database quires as well where input parameters are based either on text fields or coordinates by a click in the map. The UI of dpWebmap is following the design language Material Designed which is developed by Google in 2014.
Material Design uses grid-based layouts, responsive animations and transitions, padding, and depth effects such as lighting and shadows.
Both applications are equipped with the module Area Maintenance, which is based on a data model consist- ing of outdoor environment objects, stored and represented in the same map product. The components and tables created in the databases are all related to the outdoor environment and objects such as trees, green areas, buildings and roads are just a few out of many objects that can be represented in the map through the Area Maintenance module. Each object constructed as earlier mentioned by subtypes, attributed and components.
Organizer and Organizer Webmap are two other modules possible to add to both applications, but is not yet used by any of the housing companies, nor by any other dpMap users. Organizer is a work manage- ment module with its own data model in which projects, tasks and activities can be created, with many possibilities such as adding documents, notes, milestones and much more. By using the Oragnizer module in dpMap and the Organizer Webmap module in dpWebmap, a connection is established between the two applications, giving possibilities to send activities between field and office. Each activity is connected to a workset and only possible to post/commit from dpMap, which makes it possible to edit and manage objects in the field in a controlled way. Organizer is a flexible module and depending on the configuration, it is possible to use it as a tool for long-term projects involving many people as well as small every-day tasks.
Organizer and Organizer Webmap are two modules that could possible lead to a work flow improvement in terms of outdoor environment controls, see Figure 6.
Figure 6: A brief model of the housing companies owned GIS system. The black lines represents the system today and the blue lines the suggested additional modules for further workflow improvement. A Java client and a web and mobile client are used with the additional Area Maintenance module on both clients. The Area Maintenance module is based on a data model representing spatial objects in the outdoor environment, stored in a Oracle database. The suggested two modules, Organizer and Organizer Webmap, are used for communication between the Java client and web and mobile client for possibilities to send changes between field to the office. Both modules are working towards the same Oracle database as the rest of the system.
3 Methodology
The method carried out to achieve the stated goals in this thesis is the Contextual Design, which is a method applied to get an understanding and possible improvements of the working process in office and fieldwork.
It is a structured user-centered front-end design process that provides methods to collect data about users in field and to interpret and consolidate that data in a structured way. The data is further used to create and prototype products and service concepts, where the aim is to understand users in order to find out their fundamental intents, desires and drivers [21]. Beta testing is performed to evaluate the prototype and introduced to the users during a presentation, who are further given the possibility to explore the prototype on their own for some time. A questioner is sent to the participants with questions related to the user requirement presented in the Affinity Diagram and Consolidated Sequence Models conducted during the Contextual Design. In Section 3.1 it is explained why the Contextual Design method was chosen in terms of this case, followed by the a description of the method itself and how it was implemented in this study in Section 3.2. The prototying is further described in Section 3.3.
3.1 Choice of Method
Svenska Bost¨ader and Stockholmshem are both providing existing GIS solutions by the same GIS provider.
What will be identified along this thesis are what new functions to add, how to grow the existing GIS so- lutions and how to improve the user experience and the workflow between office and field, where a better
understanding of their work in field and needs for the users are of high importance. Evaluating suitable methods for this type of case resulted in a few relevant candidates. Since the aim is to work towards the housing companies and their needs, performing interviews with the field workers seemed like a reasonable approach in first place to gain data about the users needs. Due to the fact that this case is based on an already existing and used GIS solutions, which are about to be improved, another relevant approach could have been to perform usability tests, where products or services are evaluated and tested with representative users and is typically performed in the way that participants complete tasks while observers watch, listens and takes notes [22].
By already being briefly introduced to the housing companies working process and their layman users, interviews seemed not to be the optimal choice to achieve the stated goals of this case. Sitting in a room and trying to explain the workflow, the usage of the GIS solution and the observed problems out in the field may lead to a risk in exclusion of performed steps and experienced problems or challenges. The designer may also find it hard to get a context of the use and relevant and useful functionalities and improvements of UX may be undiscovered. Going for the usability testing would be relevant to explore obstacles and issues of the used GIS solution and to identify what encourages the user to not use it out in the field, along with those parts of the GIS solution that does works as expected. This could be useful in terms of improving and identifying useful existing functionalities and UX, but may not be as effective in exploring and gaining new ideas in terms of the workflow between field and office. However, a field-study method seemed after some consideration to be the most relevant approach. A field study is a research activity that takes place in the users context rather than in an office or lab where people are observed in their natural environment, which is suggested to use in terms of designing new products or capabilities [23]. There are plenty of field studies to choose among, where some may include interviews and some only observations. Evaluating suit- able field-case methods for this case, having both functionalities and UX in focus, the Contextual Design method is found to be well suited to achieve the stated goals. Contextual Design is a user-centered front- end step-by-step design process built upon a in-depth field research to drive innovative design. It includes techniques to analyze and present user data, drive ideation from data, design specific product solutions, and iterate those solutions with customers [24]. The aim is to understand users in their own environment, to get an insight into their lives and to finally design problem by using that insight [24].
Contextual Design is developed by H. Beyer and K. Holtzblatt and has been used since 1988. Is is a process that can be used in many ways depending on the needs of the project and be used to refine or extend existing products [24]. Multiple books have been published covering the process of the Contextual Design method, where two of these area used as a guidance during this thesis. Rapid Contextual Design: A How-to Guide to Key Techniques for User-Centered, 2005by K. Holtzblatt, J.Burns Wendell , S.Wood and M.Kaufmann is the core book used and is a guide for practitioners of the most frequently used Contextual Design techniques and contains step-by-step instruction [21]. Three Rapid Contextual Design processes are suggested in the book based on what type of problem that is about to be solved. For this thesis, the process Focus Rapid Contextual Design is applied which is the most complex processes of the three pro- cesses suggested and is according to the book the most suited rapid method for developing significant new features, web site redesign for transactions, next-generation systems or supporting a coherent task [21].
The book Contextual Design: Design for Life, 2017 by K. Holtzblatt, H.Beyer and M. Kaufmann was used to get a deeper understanding for some of the steps and techniques explained in the previous book.
In terms of the prototype evaluation, four methods have been considered; focus groups, field studies, usability testing and beta testing, where the last one was chosen to perform. Using focus groups, which is a moderate group discussion with possibilities to learn about users attitudes, beliefs, desires and reactions to concepts [25], would not give the layman user enough time to explore the prototype. Field studies and usability testing were not possible to perform in terms of the users lack of time. However, beta testing which is a method that allows the user to try and test the finished results and provide feedback by answer-
ing questionnaires [26], was suitable in terms of this case since the user gets the opportunity to explore the prototype in detail at any suitable time.
3.2 Focused Rapid Contextual Design
The Focused Rapid Contextual Design is a process that characterize the user population [21]. There are three phases to Contextual Design: Data Collection and Interpretation, Consolidation and Ideation and Detailed Design and Validation, represented in Figure 7.
Figure 7: The Focus Rapid Contextual Design Process.
The aim is to immerse in the life of individual users by field visits and interpretation of the data, this by using models to show a big picture of the whole market. The big picture is further used to drive ideation and inventing new product concepts from the user data, which are designed with concrete user interface and behaviour, validated and iterated with users [27]. The Focus Rapid Contextual Design process consists of the following top-level steps: Contextual Inquiry, Interpretation, Affinity Diagram, Data Consolidation, Visioning, Storyboarding and Prototyping [21]. Each step is briefly explained in the following sections.
3.2.1 Contextual Inquiry
The Contextual Inquiry, explained in detail in [21] and [27] in chapter 4 and 3 respective, is a one-to-one field data gathering technique and guides the designer in going out into the field to talk with and observe people while they work. Contextual Inquiry immerses designers in the users whole life - including those aspects which the user does not know how to articulate. Each Contextual Design projects starts with a project focuswhere the aim is to define what the designer needs to pay attention to. What problems that needs to be solved, who are the users that are affected, which of the users activities and tasks are important as well as the scenarios and locations that are relevant. All this needs to be defined in the beginning of a process and will later be used to guide the interviews. The Contextual Interview is based on four principles:
context, partnership, interpretation and focus, where each principle defines an aspect of the interaction and are used guide how to run the interview. The principles are further described below.
Context
The aim is to get as close to the activity as possible to see how the target activities fit into time and place, what platforms, products, or devices are used, how people collaborate or coordinate to get things done and how policy or organizational structure affects what people are doing.
Partnership
The work with users are performed as partners where the user leads the designer through the activities while the designer makes observations and asks questions. Questions are not planned beforehand, rather the project focus is used to guide the conversations.
Interpretation
The aim with the observations in the field are not just to collect the facts of what the users are doing, but also to come back with accurate interpretations. The aim is to create a shared understanding of the users actions by sharing hypothesis with the user.
Focus
Steers the conversation and tells both the designer and the user what to pay attention to.
The Contextual Inquiry was performed with both housing companies with one garden coordinator par- ticipating from each housing company. A general interview was first held at the housing companies office followed by a field interview that was performed in an outdoor environment of a housing area.
3.2.2 Interpretation Session
An Interpretation Session occurs within 48 hours after the field interview where the intention is to under- stand the field interview data. During the Interpretation Session, two or more participants are gathered to capture insights and key issues while carefully listening to the details of the interview told by the inter- viewer. The captured insights and key issues, also referred to Affinity Notes, are later built in an Affinity Diagram which are further used in the design. During the Interpretation Session, the participants are asking questions, draw out details that the interviewer might have missed and indicate what is important to capture in the Affinity Notes. To deliver the best results, the participants should represent a range of job functions and viewpoints. During the Interpretation Session it is important to state what role each participants has, so that the person knows what they should mainly focus on.
Besides the capturing of Affinity Notes, Work Models are created during the Interpretation Session. A Work Model is a physical diagram that captures complex qualitative data and the structure of the users work or activity, which helps the participants and designer to see the work structure. There are five Con- textual Design Work Models; Flow, Sequence, Cultural, Artifact and Physical, where the Sequence Work Model is the key model used for Focused Rapid Contextual Design, which is a model capturing the steps a user takes to perform a task. Each new task is represented by a sequence composed by the steps the user takes to perform the task, the trigger that causes a user to start a new task/particular step and the intent - the reason the user is doing the task or the step. The level of detail of the sequences depends on the project, where for the focus of the user interface or usability of a tool, it is recommended to capture the sequences at the level of the clicks, meaning any time the user presses a button or clicks and inputs something to the user interface.
For the Focus Rapid Contextual Design, it is also recommended to capture the Physical and Artifact Model.
The Physical Model draws the context and the Artifact Model gives an overlook what things and tools the users use in their work. After the Consolidation performed in Section 3.2.3, the Work Models and the Affinity Diagram will represent the physical representation of the characteristics of the user population.
Further information about the Interpretation session and the Working Models can be found in [21], Chap- ters 5 and 6 and in [27], Chapters 4 and 5.
The Interpretation Session was as suggested accomplished within 48 hours after the Contextual Inquiry during the case study. One more participant, besides the interviewer, participated during the session - a GIS consultant at Digpro AB with major focus on their web GIS solutions. The interpretation roles were spread between these two and the interview was gone through twice; once to make the Affinity Notes and once to draw the Sequence Models. The Artifact Models were created meanwhile by defining the used artifacts during the fieldwork and the Physical Models was chosen to not include due to only two locations involved.
3.2.3 Affinity Diagram and Data Consolidation
The Affinity Diagram and the Consolidated Sequence Models represent all issues and patterns of work across the user population. An Affinity Diagram is constructed by using the Affinity Notes captured during the Interpretation Session and group them according to the similarities of their contents. Each group is labeled with a colored note which represents a distinct level in a hierarchy. The groups are then combined with other groups and ends up in diagram of observations in a hierarchy of up to three levels. The Affinity Diagram acts as the voice of the customers/users and shows the common issues, themes and scopes of the customers problems. The Consolidated Sequence Modeling is based on a similar process to the Affinity Diagram process. The Individual Work Models created during the Interpretation Session, representing par- ticular examples of how the work hangs together for one person, are grouped and consolidated according to their contents and similarities to reveal the structure of the work across users. The goals of Sequence Consolidation is to produce one model for each of the primary tasks and the steps a user performs during these tasks. Further information about the Affinity Diagram and Data Consolidation processes are found in [21], Chapters 7 and 8 and in [27], Chapters 6 and 8.
In terms of the case study, this process was split in two parts; the first part with focus on the Affinity Diagram which was performed with the interviewer and the GIS consultant introduced in Section 3.2.2 and the second part with focus on the Sequence Model Consolidation performed by only the interviewer. This ended up in three Consolidated Sequences and a two-level Affinity Diagram.
3.2.4 The Wall Walk and Visioning
The goal with the Wall Walk and the Visioning is to cover both divergent and convergent thinking, this by first come up with multiple possibilities and finally settling on a mutually acceptable approach. Doing so, it will reduce the risk to limit the creativity by converging too quickly on a single idea without considering other options. During the Wall Walk, which supports the divergent thinking and is the step right before the Visioning, the participants explores and becomes familiar with the data by reading the Affinity Diagram and Consolidated Sequences, individually and in silence while writing down their design ideas. The par- ticipants are then gathered together to create two lists; one with key issues and one with hot ideas. The key issues represents the top user needs experienced by walking the wall, whereas hot ideas are big design ideas, seen as starting points of further ideation.
The Visioning, which supports the convergent thinking and is done right after the Wall Walk, is a group storytelling process. One person starts to tell a story from the users point of view, performing the tasks which where explored during the Contextual Inquiry, with the new technology and process that are about to be invented. Each story starts with a hot idea and during the story telling, the other participants are adding to the story as it unfolds. Meanwhile, one of the participants are hand-drawing the visioning. The aim is to create three to four individual visions, evaluate each of them, identify what works and doesn’t work and finally take the good parts of each vision and create a final consolidated vision which will be used to guide the work further. To encourage to think more systematically about the redesign and include possible changes in technology, user interface function, underlying business rules, role definition and process, it is important that the visioning team includes people who understands the technology, design and work prac- tice. Further information about these two methods are found in [21] and [27], Chapters 10 and 11.
The Wall Walk and Visioning were performed in a group of four during the case study, including three GIS consultants from Digpro, each one having expertises within different areas, and the interviewer. Due to time restrictions during the Visioning, it was not possible to draw any Visions in place as recommended - the ideas and stories of the participants were rather written down during the session and drawn by the interviewer afterwards.
3.2.5 Storyboarding
Storyboarding, which is explained more in detail in [21] and [27] in Chapters 12 and 13 respective, is a process to work out the detail of how to support the sequences within the new system. It can be by intro- ducing a new way to accomplish an activity or by doing it in the old way. The Storyboarding is based on the Sequence Models and the Affinity Diagram and makes sure that the intents and steps that are critical to the work are considered. Only the relevant sequences are storyboarded. During the Storyboarding pro- cess, the participants are drawing step-by-step pictures of how people will work in their new world, where the sequence models represents the core tasks. The drawing includes manual steps, rough user interface components, system activities and automation and documentation uses. Each interaction with the system is represented by a step in the storyboard.
The Storyboarding was performed in two steps, involving the interviewer and the GIS consultant presented in Section 3.2.2. During the firs step, the web GIS solution already used by the housing companies were inspected carefully with focus on the already existing functionalities that did fulfill the user requirements represented in the Affinity Diagram and Consolidated Sequence Models. These functions were decided to keep, some with minor changes in mind. Further on, new functionalities were discussed with the UX design in mind, that could fulfill the remaining requires or to solve the key issues discovered during the Contextual Inquiry. Functionalities of other web GIS solutions provided by Digpro were looked over to see what functionalities that could be implement straight away, and what functionalities that would need further development. During the second part of the process, three storyboards were created, each based on one sequence where each frame represents one move for the user and presents functionalities and UX briefly.
3.3 Prototyping
The final step in Contextual Design is the prototyping process, which is an iterative process where the aim is to test the design concepts and clarify the function. Usually a paper prototype is constructed, which is a paper representation of the product with movable parts, and used to test it with the users in field inter- views. The users interacts with the prototype during the interviews and makes modifications while they perform different tasks. This is followed up by an Interpretation Session to capture key issues and new functionalities that are further implemented to improve the prototype. The prototype is redesigned after each round of interviews and tested again in the next round. This iterative process ensures that the final functions and user interaction layout suits the users. Prototypes act as a bridge for communicating between user and designer, where the users get a chance to interact with the product just like they would with any other product, and by using their feedback it is possible to improve the design. Further information about prototyping is found in [21] in Chapter 12 and 13 and in [27] Chapters 16 and 17.
In case of this thesis, a prototype is configured in the housing companies already existing web GIS solu- tions, this to give them a chance to get ever closer to the finished product. The prototype was not presented in paper form as suggested by the Contextual Design method. This choice was made due to the fact that this study aims to improve an existing GIS solution and not to develop a new one. The realization of the prototype was based on the web and mobile client dpWebmap with the additional Area Maintenance, Or- ganizer and Organizer Webmap modules, and was configured according to the Storyboards resulting from the Contextual Design. Some adaption were made in the housing companies data structures and specific
