AGILE USER EXPERIENCE Tina Øvad Radiometer Medical ApS and Aalborg University tina.oevad.pedersen@radiometer.dk ABSTRACT This paper describes a PhD project, exploring the opportunities of integrating the processes and methods of user experience design and agile software development, particular in the development of medical devices. First I briefly introduce the background for the study and the different attempts having been done in the past to make this integration. Secondly, one research hypothesis and five research questions are formulated and I propose different methodologies to investigate these questions. Thirdly, I describe the initial work done, consisting of interviews and a literature review. Finally, the future work of the thesis is introduced. KEYWORDS Agile user experience, PhD thesis abstract. 1. INTRODUCTION The aim of this work is to explore the opportunities of integrating the processes and methods of user experience design and agile software development. I initially argue that an integration is possible seen from a theoretical point of view (DeMarco Brown, 2013; Kollmann, 2008) and attempts of having the two frameworks integrate have been made by i.a. Beyer et al., (2004) Kollmann (2008); Miller (2005) Singh (2008) and Sy (2007). In addition interviews conducted by me revealed that Danish companies are very interested in such an integration (Øvad et al., 2014). Advantages of such an integration in terms of reducing wrong design decisions (Memmel et al., 2007), secure better quality of the user experience (Sy, 2007) and make an improved product (Meszaros and Aston, 2006; Singh, 2008). The thesis will therefore investigate how to make an integration of the two by means of a user experience toolbox suited for development in an agile software development environment. Since the project is made in collaboration with a company developing medical devices, the research pursued throughout this project will in particular examine the development of a user experience toolbox suited for the development of medical devices. This focus is taken since there has been an increasing focus on user experience in medical devices, especially in the last five years and both ISO 1 and FDA 2 have published documents on human factors guidelines and standards. However, there are still many unsolved challenges, not least the proliferation of sophisticated medical equipment from specialized labs into more widespread use. This induces a change in the user base - from trained and experienced operators to regular hospital staffs such as nurses, who are expected to operate the devices, perhaps not even on a regular basis. Furthermore, the emerging standards do not in fact specify how the desired user experience should be achieved. As a result of the increasing focus on user experience on medical devices, it is desirable to have the user experience design permeate the whole software development process in order to make sure that these guidelines and standards are followed and documented. Since many software development projects nowadays are developed within an agile environment, it is necessary to investigate how best to integrate user experience design in an agile software development, so user experience design may become a natural part of the software development process. 1 International Organization for Standardization 2 U.S. Food and Drug Administration
2. RELATED WORK As mentioned, attempts have been made to integrate user experience processes (e.g. front loading and user tests) and methods (focus groups and prototyping) in the agile processes (e.g. Scrum and XP). One solution has been to have the user experience designers run sprints ahead of the rest of the development group, by having a phase 0 or sprint 0 (Beyer et al., 2004; Kollmann, 2008; Miller, 2005; Sy, 2007). This framework is a good transition model, but it is not a permanent solution, since the development will never have the benefits of the cross functional synergy of different professions working together, their focus are at different stages of the development process. Another solution has been to have a user experience specialist working as a satellite on the development team. The idea is that this person is supported by other user experience designers outside the development department (Kollmann, 2008). A serious problem with this approach is that if the satellite person is disconnected from the user experience team, the results from the research and the tests are disconnected from the whole user experience vision (Kollmann, 2008). A third approach used when working with Scrum as the chosen development framework, is to have the role of product owner assigned to two peers, where one represents the traditional role with focus on traditional functions and the other is focusing on usability and user experience. The usability product owner is responsible for establishing the user experience vision for the product together with different personas (Singh, 2008). The UScrum approach is well-suited for products which are novel and complex. If a topic is well-known, the advantage is not big enough in relation to having the coordination overhead in having two product owners (Singh, 2008). Neither of the attempts have succeeded in doing a full integration, meaning that the agile development and the user experience design are done synchronously, resulting in optimal cross functional synergy. Furthermore, there exists little guidance on how to integrate them in practice and the day-to-day work in which the processes are used are uncharted (Ferreira, 2012). The three main findings emerged from the previous attempts are: (1) The user experience designers initiate their work before the software development team is engaged, inducing that the development do not achieve the cross functional synergy of working interdisciplinary. (2) The software developers have not been used as a user experience work resource. (3) Guidance is missing on how to integrate the two processes in the day-today work. 3. RESEARCH GOALS AND METHODS As pointed out above, three main findings were detected and I suggest one possible solution is to have the software developers do some of the user experience work themselves. In this way the user experience will permeate the whole development and the user experience practitioners have more resources to focus on the user experience vision and make extensive user tests. To guide the software developers on how to make user experience work, a user experience toolbox is to be developed and this leads to the research hypothesis: To what extent can a user experience toolbox, developed to be used by software developers in an agile software environment, facilitate synchronous work with agile development and user experience design? To investigate this question five research questions are put forward: What is the current state of user experience and agile processes within Danish companies? Which user experience methods, procedures and tools are suitable when developing medical devices in an agile development environment? Is it possible to tailor a user experience process, suitable for the development of medical devices in an agile environment? To what extent can the software developers be directly involved in the UX processes? How can this toolbox be integrated in an agile software development process?
To answer these questions the project is divided into two phases (A and B). Phase A is to investigate how Danish companies currently work with user experience methodologies in an agile development environment, this is done by a phenomenological study. Furthermore, previous attempts of having user experience methodologies integrated in software development processes are reviewed and processes, methods, procedures and tools suitable for the development of medical devices in an agile environment are identified. The result of phase A is a framework in which different user experience processes are stated, together with different user experience methodologies, both suited for the development of medical devices in an agile software development. Phase B is an iterative process between experimental evaluation and analysis of the components of the framework found in phase A. Phase B is therefore split into two parts; B1, which is more experimental and B2, which is more analytical. B1 and B2 will be repeated a number of times. In part B1 processes and methods from the framework are applied experimentally in the agile software development processes at a company developing medical devices. Since it is an iterative process, individual processes and methods are validated one at a time. In part B2, the findings from B1 are analyzed and conclusions about the suitability are made. The results will be that the applied processes or methods are either; discarded, accepted or modified to suit the context of development of medical devices in an agile process. In the latter cases, the processes and methods are collected to a user experience toolbox and the next iteration is initiated. 4. PRELIMINARY RESULTS Interviews with nine interviewees from eight Danish companies have been conducted and documented in Øvad et al. (2014). The interviews aimed at getting an insight in how Danish companies currently work with usability and user experience design in an agile development environment. This investigation was done as an ethnographical study and covered the research question: What is the current state of user experience and agile processes within Danish companies? Via the interviews I have identified two typical organization types when working with user experience design. Type I is larger organizations with a specialized user experience department (or team). In these organizations user experience specialists can be called upon e.g. to carry out user studies when necessary or relevant. Type II is smaller organizations, with no user experience specialists and no resources to build such a department (or team). The difference between the two types of organizations induces that the potential integration of UX processes into the agile development can be done in different ways, illustrated in figure 1: Type I: UX Team Software Team Sprints UX Team UX Team Type II: Software Team + UX Team Sprints Figure 1. The typical organization types when working with user experience design in agile software development environments (UX = user experience)
In addition it was found that there is a major lack in the processes of working with user experience design. This indicates that before making an integration of user experience design and agile development, user experience processes are to be described and documented. Furthermore, the interviews revealed that the companies have taken agile software development to heart especially Scrum, and the companies seem to adhere to it. This is an excellent argument for a beneficial integration between Scrum and user experience design. By this integration user experience design can gain some of the benefits the software development has gained from this framework; more transparent work, always something to show to the customers etc. (Øvad et al., 2014). Furthermore, a literature study has been conducted of the current state within the research field. The literature review was done to generate ideas for developing user experience processes and to collect methods suited for the user experience toolbox and covered the research question: Which methods, procedures and tools for experimental verification of user experience suitable for medical devices in an agile development context? Different methods, procedures and tools were identified as being suitable and are now to be finally evaluated as being suitable for developing medical devices in an agile software environment. 5. CONCLUSIONS This paper introduced my PhD project investigating which user experience processes, methods and tools are suitable for integration into an agile software development environment with a focus on developing medical devices. One research hypothesis and five research questions relevant for the field of integrating user experience design and agile software development have been formulated together with methods on how to answer them. The following paragraphs conclude the presentation of the thesis by firstly summarizing the thesis scientific contribution and secondly describing the future work being planned. 5.1 Contribution Summary By answering the research hypothesis and the five research questions, the project will provide the scientific community with insights on key aspects on how to do user experience design in an agile software development environment, especially when developing medical devices. The challenges of having user experience work and agile software development run synchronously will be investigated within a real software department. The expected outcome of the project is a described and documented integration of user experience design and agile software development. This is done by means of the user experience toolbox, which contains: (1) Described and documented user experience processes suited for the development of medical devices in an agile development. (2) A description of different user experience methods, including the effects of them, the load of using them and the data generated from them. By developing the user experience toolbox, it is possible for the software developers to make some of the user experience work themselves, resulting in user experience permeating the development process. This entails that the development follows the human factors guidelines and standards put forward by ISO and FDA. Furthermore, time is allocated from the user experience designers, inducing more time to make initial user experience investigations prior to the software development work and to keep track and maintain the user experience vision of the medical device. 5.2 Future Work The developed user experience processes and the collected processes and methods are to be tested and further developed in close collaboration with the software developers at a medical device company. Currently I am designing workshops for the software developers on fundamental methods, such as: Focus groups
Contextual Inquiry Low-fi prototypes and tests Cognitive walkthroughs More methods are to be added to the list when the project progress I am to teach the software developers on how to apply the different processes and methods. The idea is that the developers are to be interviewed before the workshop about their expectations to the workshop, then taught in the current process or method. After the workshop the developers are interviewed again to hear their thoughts about the process or method. Finally, they are interviewed after they have used the process or method, in order to record their experiences and their input to the use of this process or method in an agile software development environment. Then the process or method is either discarded, accepted or modified to suit the context of development of medical devices in an agile process (as mentioned in section 3). In addition, a visit to an American company producing electronic test tools and software is planned. This visit is to gain input to the user experience design process and to investigate how a company work with user experience design in an agile development environment, when the company is very design driven. Furthermore, a longer stay at another American company is to be confirmed. The purpose of this stay is to have the opportunity to implement and validate the user experience toolbox in another company. I hope by participating in the doctoral consortium, to have inputs and ideas to different user experience process and methods to be implemented in the user experience toolbox. Furthermore, I hope to have inputs, ideas and inspiration on how to convey the idea of working with user experience design to the software developers. REFERENCES Beyer, H., Holtzblatt, K., Baker, L., 2004. An Agile User-Centered Method: Rapid Contextual Design. DeMarco Brown, D., 2013. Agile User Experience Design - a practitioner s guide to making it work, First. ed. Morgan Kaufmann/Elsevier. Ferreira, J., 2012. Agile Development and UX Design: Towards Understanding Work Cultures to Support Integration, in: Workshops. Presented at the CAiSE 2012, Springer-Verlag Berlin Heidelberg, pp. 608 615. Kollmann, J., 2008. Designing the User Experience in an Agile Context. Faculty of Life Science, University College, London, London. Memmel, T., Gundelsweiler, F., Reiterer, H., 2007. Agile Human-Centered Software Engineering, in: People and Computers XXI HCI but Not as We Know It. Presented at the HCI 2007, British Computer Society. Meszaros, G., Aston, J., 2006. Adding Usability Testing to an Agile Project. Presented at the AGILE 2006, IEEE Computer Society. Miller, L., 2005. Case Study of Customer Input For a Successful Product. Singh, M., 2008. U-SCRUM: An Agile Methodology for Promoting Usability. Presented at the Agile 2008 Conference, IEEE Computer Society. Sy, D., 2007. Adapting Usability Investigations for Agile User-centered Design. JUS - J. Usability Stud. Vol. 2, 112 132. Øvad, T., Larsen, L.B., Yndgaard Sørensen, S., Skriver, J., 2014. The Current State of Agile UX in the Danish Industry. Journal of Usability Studies. (Pending approval)