To summarize, a representation technique is needed that is expressive in terms of detailed design, sketchy in its expression, and versatile in its ability to create fictions.
I have experimented with a representation technique that seems to fulfill these three criteria quite well. Animated use sketches are animated movies expressing important scenarios in the intended use of the future artifact. The movies are produced using simplified and rapid cutout animation techniques with the intention of conveying a sketchy and unpolished impression. What follows is a case study where animated use sketches were used. The technique was assessed informally by the stakeholders in the design process. I present the assessment results and then discuss animated use sketches in relation to existing representation techniques and design representation in general. The concluding remarks address the cost-benefit tradeoff of animated use sketches.
The industrial company Alstom Power Sweden, which I will refer to simply as Alstom, manufactures and sells filters for cleaning exhaust fumes from heating plants, paper mills and similar industries. In this context, a filter is typically a 10x10x20-meter assembly of heavy electromechanical components. Each filter is built to fit the cleaning requirements and specific properties of the industrial plant in question, albeit drawing on general principles and best practices. The filter has to be inspected at regular intervals and some parts may need to be repaired or exchanged. Alstom has service contracts with many of its customers, specifying annual inspections and other maintenance work.
Performing an annual inspection involves climbing into and crawling around a dark and dirty filter in order to assess the status of roughly a 100 predefined inspection points. The inspector has to make his way to each inspection point inside the filter, assess its condition; then, from the entrance hatch, report the status to an assistant waiting outside the filter with a paper form on a clipboard. When all inspection points are complete, the inspector disengages from the filter, cleans himself up, travels back to his office, reworks the inspection notes into a MS Word report and sends it to the customer. He then negotiates the maintenance plan with the customer at a later time.
Alstom's goal was to find a way to simplify inspection data entry, report generation and maintenance negotiation. Ideally, the process would be completed before the inspector leaves the customer site.
A project team was formed involving employees from relevant departments at Alstom, as well as system developers from Sikroma, a CAD company aiming at developing new software products for structured- information management in the mechanical industry. I was the interaction designer of the team. A collaboration agreement within an existing local R&D consortium ensured a high level of trust and openness among the team members.
Based on initial information from Alstom, I developed a plan for supporting knowledge-management and coordination among maintenance workers. However, a workshop with experienced maintenance staff indicated that the problems of inspection data and report generation, as previously described, were more relevant.
I sketched a concept based on voice interaction with a wearable device for data entry during inspection, followed by a wirelessor "dock"transfer to a dedicated laptop application for data cleanup and report generation at the customer site. The inspection data would then be synchronized with a central maintenance database. The new concept was presented to Alstom in a series of hand-drawn sketches and text explanations in a six-page document that I presented to Alstom along with several specific questions. The maintenance staff indicated that the concept was promising but needed some revision.
I revised the design based on the feedback and created a two-minute animated use-sketch to express it. Sikroma developed a simple demonstrator using implementation-strength database technology to explore the technical feasibility of the concept. The core functions of the demonstrator included database storage, laptop data management, and report generation. Voice interaction was simulated using typed text input and output.
The animated use-sketch and the technical demonstrator formed the basis for a concept milestone workshop with maintenance and IT support staff from Alstom. The general outcome was positive and the concept was judged as (1) clearly useful for the inspector, and (2) probably efficient in terms of saving time (and hence money) for the maintenance department. Following the workshop, Alstom resolved to explore the internal financial and political conditions for developing the concept into a working tool. Sikroma plans to use the results as a basis for new products to support integrated maintenance of mechanical constructions.
The animated use-sketch was created rather late in the concept design. It was intended as a vehicle for communication at a milestone stage, when the client would be approaching a decision as to whether the idea merited further investment of development time or not. My goal was to balance the presentation in a workshop setting with openness to reactions, comments, and questions rather than making a one-sided sales-pitch.
The "movie," which was composed of backgrounds based on public-domain visual elements (most of them retrieved from old books in the "industry" section of the public library) and hand-drawn foreground elements, was built in a simple style. The animations were rough and non-anatomic, most closely similar to cutout animation. Background sound effects were drawn from public-domain sources. Parts of the dialogue were custom-recorded to illustrate the central mechanics: the voice interaction during inspection. Simple sound-processing techniques were applied to make the system's spoken output sound more computer-like. There was some spoken dialogue that did not concern the core idea of the concept; for these I used nonsense language. The animation technique and style was chosen to illustrate the work environment and use sequences of filter maintenance, while not coming across as too finished, polished or sales-oriented.
It took me 25 hours to create the animation, including the development of a storyboard (five hours), preparation of image and sound assets (12 hours), animation, editing, titling and rendering (eight hours). I used familiar tools: Adobe Photoshop, CoolEdit, Macromedia Director and Cleanerwhich means that not much time was spent on learning the tools or solving technical problems.
In the case described above, the animation technique was assessed informally through structured group discussions with stakeholders in the design process: maintenance and IT-support departments at Alstom. The animated use sketch was perceived to provide a strong grasp of the design idea and how it would work in practice, much better than a paper presentation or a conventional slide show. It offered, in the words of a participant, an "overview of what could be done and how it could be done" if the new concept were implemented. As the animated use sketch was introduced before the technical demonstrator, it served as an advance organizer that supported an understanding of the level of technical detail presented in the demonstrator.
Specifically, participants found the animation style to be clearly different from advertising and sales material. It appeared tentative and they felt comfortable in asking questions and pointing out weaknesses in the presented design concept. They noted the difference from sales-pitch situations, which they described as taking a chance on an investment where the implications were not fully understood.
The participants found two drawbacks with the animated use-sketch as a representation technique. First, they found the use of nonsense voices in non-critical parts of the story confusing. It would have been better to use the same voices throughout and merely have them recite non-specific lines. Second, as a participant pointed out, it is easier to remember a week later what was said in a meeting when it is documented on paper. To address this point, a handout could have been prepared in the form of a movie storyboard with key frames from the animation and key dialogue typed next to the frames.
The creation of design representations in general serves at least three purposes. First, it creates a vehicle for much of the thinking involved in design. Sketching in various media is designing, rather than merely packaging a design that already exists in the head. Ideas develop in creating the representations. Second, it facilitates communication and collaboration. Most interaction design involves more than one actor; representations, therefore, are essential in driving the work forward. Finally, design representations serve a rhetorical purpose. An attractive representation can be instrumental in persuading a client.
The question of design representation in interaction design is complicated by the nature of our design material. When the object of a design process is a relatively static artifact presented on a screen, traditional representation techniques such as pencil sketches, storyboards and prototypes in program code work well. However, we are increasingly facing design situations where the intended use takes place on the move, using various mobile and embedded interaction surfaces. Another complication for the designer occurs when the essential use qualities of the intended design depend on highly interactive or complex behaviors which are costly or difficult to express in conventional prototyping techniques. In situations like these, temporal and narrative representation techniques are needed. Text scenarios are a standard technique; play-acting and improvisation are increasingly used for similar purposes [7, 3]. Storytelling, using moving images, is another narrative technique, one which offers the possibility to express a relatively detailed view of the intended design.
Design representations where use stories are told in moving images, such as film or video, tend to communicate and persuade rather than merely illustrate. Moreover, it takes some time and effort to create moving-image representations, which makes them less than ideal for rapid thinking-by-sketching. Even video prototypes, which are intended to serve as sketches more than anything else, can be awkward when used as sketching tools [8, 9]. It's possible to speculate that the overhead involved in transforming ideas to video is enough to break up the smooth conversation between thought and sketch ; however, a design team that practices video prototyping may reach a level of proficiency that is not otherwise possible.
Our previous experience in creating video use sketches seems to indicate that such representations run the risk of being interpreted as rhetorical and persuasive [2, 4]. Classical video use sketches such as the Apple Knowledge Navigator or Starfire from Sun suffered from similar reception. The viewers tend to "lean back" and react to the video as a persuasive wholewhether they like or dislike the presented ideabut it typically requires significant work to engage the viewers in a more detailed and constructive discussion of the design concept.
An animated use sketch is an attempt to use moving-image storytelling in interaction design not only for persuasion but for constructive communication. To this end, animation is preferable to video since it can be made to look less definitive. The collage style suggests the accidental nature of the visual material used as backgrounds. Characters and objects in the figure plane are cartoon-like. Animation is jerky and non-anatomic with few key frames and no interpolation between them. The informal assessment indicated that our approach worked as intended. The stakeholders in the design process felt that the animated use sketch conveyed the design idea more effectively than a conventional paper and slide-show presentation, and they felt comfortable in asking questions and pointing out weaknesses in the proposed design.
Was it worth spending 25 hours on a two-minute animation? After all, a proper document specifying the design concept in detail could have been prepared in less than 10 hours.
It's significant that it took me some time to find a style that seemed appropriate, and that the visuals may have been slightly overworked. It could have been possible to cut the production time by using simpler drawings and backgrounds, or by hiring a more skilled artist. (Using simple computer-drawn foreground elementswhich is certainly faster than hand drawing followed by scanning, cleaning and coloringis probably not a good option, however. It would make the expression less tentative and possibly violate the intentions behind the technique.) Either way, it is safe to assume that the next animated use sketch I produce will take less time now that an appropriate style has been found. A production cost on the order of 15 to 17 hours is reasonable to expect for a comparable result.
More importantly, though, the animation proved effective in conveying to the clients a clear understanding of the design concept and its use qualities. My experience, as well as studies by other researchers, indicates that this outcome is hard to achieve by means of conventional specifications and documents . Spending a few hours of designer time to improve mutual understanding at a concept decision-point may prove to be a sound investment further along the development process.
1. Att inspektera ett elfilter. [Inspecting an electrical filter.] Produced by Viggen CMC, Växjö, Sweden, 2003. Low-resolution version available at http://webzone.k3.mah.se/k3jolo/ Material/sitepAlstomQ.mov
2. Andersson, O., Cacciatore, E., Löwgren, J. and Lundin, T. (2002). Post-hoc worknotes: A concept demonstration of video content management. Proc. 10th ACM Int. Conf. Multimedia (MM02), pp. 670-71. New York: ACM Press. Video and paper submission.
4. Löwgren, J., Bonneau, J., Möller, A., Rudström, ö., and Waern, A. (2000). News on demand considered useless: An explorative assessment of database news publication features. In Building tomorrow today (Proc. i3 Annual Conference), pp. 19-25. [Available at http://webzone.k3.mah.se/k3jolo/Download/nod0006.pdf]
5. Mackay, W., Ratzer, A., and Janecek, P. (2000). Video artifacts for design: Bridging the gap between abstraction and detail. Proc. Designing Interactive Systems (DIS `00), pp. 72-82. New York: ACM Press.
See also the video prototyping section of the User-Centred Requirements Handbook at http://www.ejeisa.com/nectar/respect/5.3/417.htm
School of Arts and Communication, Malmö University
©2004 ACM 1072-5220/04/1100 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2004 ACM, Inc.