MediaLAB Amsterdam is now the Digital Society School! You are viewing an archive of MediaLAB projects. Content on this website may be outdated, and is without guarantee.

Improving project work of lab-based Multidisciplinary Design Teams using Scrum & Design Thinking

At MediaLAB Amsterdam, we are currently experimenting with new ways to optimise the way the multidisciplinary (student) teams work on their projects in the lab. In order to do so, we have recently added scrum as agile work-form to the context of our regular human centered design/ Design Thinking process. Already after two months, this is giving us interesting insights and results, but in parallel presenting us with new challenges. I will share some of our early insights in this post.

Multidisciplinary Design Teams

In recent years, (design) academies and universities have been eagerly trying to adapt to and catch up with working trends in industry and commerce, where multidisciplinary team-based working is becoming prevalent. This is partly based on indications that, if performed and managed correctly, multidisciplinary team-based working appears to bring better results in terms of speed and quality of the product and the quality of interpersonal working relationships.

Multidisciplinary Design Teams (MDTs) bring together people with a range of expertise to a (design) project [1]. Multidisciplinary involves multiple sources of knowledge, skills, and attitudes towards learning and solving problems. One of the most well known design processes that MDTs follow is currently that of human-centered design.

Modern human-centered design is generally recognized to have originated at IBM in the 1980s, emphasizing four “critical steps”: 1) “Early focus upon the characteristics and needs of the intended user population”,” 2) users as part of the design team, 3) empirical and experimental measurement, and 4) iterative practices [2]. Human-centered design typically follows a linear process, which resembles a process that is sometimes referred to as “waterfall”, where insights and requirements need to be defined in great detail before design and prototyping can take place. Within that final design and prototyping phase, an iterative approach is taken to optimize the design. Human-centered design methodology can be effective in supporting MDTs to learn from design; therefore it also seems useful for educational settings.

In design education, multidisciplinary team-based working has gained more recognition in recent years, and so-called lab-based education has increased steadily in the course of the past years. Students are placed in MDTs and have an appointed space within a ‘lab’ where they often follow a human-centered design process in order to develop a design solution for a given problem.

Combining Scrum with Design Thinking

Both with Scrum and Design Thinking, there is the potential to run projects indefinitely. With Scrum, new requirements can be added to the project back log and integrated in a next sprint. In Design Thinking processes, endless iterations for improvements are also an option.

In previous semesters, we used to work with a typical ‘waterfall based’ model where through Research (and understanding) concepts were Created in the form of prototypes and finally, a final Design was worked out in more detail (see figure below).


Fig.1: RCD process, a human-centered design approach

The length of each phase was fixed, although they did vary between the phases (from 4 to 8 weeks).

In order to find an answer to the question whether this approach and the program at the time supported the MDTs in the lab in the best way, and optimises learning from design and development of collective design competencies, an evaluation study was performed among the participating students just after the semester finished. In the paper “From Clutter to Butter” (INTED2014) results from this evaluation study, are discussed  in the light of the importance of individual/ team learning, motivation and learning transfer, in supporting MDTs and the design process, and introduced the ingredients of a new, improved model, based on these outcomes.

The result showed (a.o) that:

  • The fixed structure and length of the phases created fluctuations in motivation that obstructed creativity.
  • Making the connection between the phases and transferring knowledge effectively between them, was not properly facilitated or supported within the lab.
  • Individual and team learning levels could be improved.
  • Prototyping and building ‘tangible’ concepts that could be tested and evaluated was postponed to almost in the final weeks before the deadline of a final deliverable.

On the basis of these initial insights, a new design process (see figure below) was presented, where Scrum had a role. For the improved process, Scrum promised to provide structure and the freedom for an MDT and its members to direct their own path. Additionally, the sprints and Scrum rituals can engage members to work more effectively and transfer knowledge with more ease between phases. To avoid becoming a victim of structure, MDTs should be able to manage the before mentioned length and pace of each sprint. Furthermore, the built in moments of reflection and feedback within the Scrum framework will help to foster continuous reflection.


Fig.2: The Design Cycle – improved design process for MDTs in labs

After working with this new approach for almost two months, we evaluated it and found a couple of effects that made us change the initial model somewhat:

  • It is important to have the backlog/ requirements set before starting a particular sprint. However, initial research may be performed beforehand.
  • There is no sprint (or iteration) where the Research/ Translation/ Creation parts are equally long, or equally important.
  • In the Scrum framework, it is hard integrating bigger questions or vague ideas as they are difficult to narrow down as specific tasks.

In the following figure you can see the changes:

Screen Shot 2014-03-29 at 10.49.12 PM

Fig.3: The adapted design process, integrated within Scrum

Another important change was that each team could choose a specific ‘type of sprint’ while planning the sprint. Will it make more sense to focus on research, translation or on creation and prototyping? In a research focused sprint, there is lots of room for investigation/ research, and little time for translation and creation. The end deliverable could be a research output, or a naive design based on the research. In a translation focused sprint, it is more about the understanding of insights, the team spends most time on translating the insights into requirements. In a creation focused sprint, there is almost only room for prototyping, building, trying, etc.

Screen Shot 2014-03-29 at 10.49.25 PM

Fig.4: The sprint types

Dealing with creativity and reframing goals in Scrum

Another challenge that the teams are currently phasing is that of having the chance to reframe initial goals during the sprint. Sometimes the design process urges the team to do so, but it is hard to integrate in Scrum. To facilitate this, teams can choose to run shorter sprints, so that the iterations become shorter and there is more space for exploration and therefore creativity.

INTED2014 paper

Read the paper From Clutter to Butter: Ingredients for Improving Multidisciplinary Team-based Design Education in a Lab Context that was presented at INTED2014, Valencia, Spain.


[1]  Evers, M. (2004). Learning from Design: facilitating multidisciplinary design teams. Delft: Eburon.

[2]  Friess, E. (2010). The Sword of Data: Does Human-Centered Design Fulfill Its Rhetorical Responsibility? Cambridge, Massachusetts: UNT Digital Library. Accessed February 1, 2014.

Visuals designed by Gabriele Colombo