• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!


Week 8

Page history last edited by Nina Khosla 14 years, 1 month ago

Comments (12)

Andrew Hershberger said

at 11:17 am on May 21, 2010

In "Designing Adaptive Robotic Services", Lee and Forlizzi suggest a modification to the service blueprint to show how a service adapts to users over time. Unlike the original components of the blueprint, which depict a detailed process for how the service will work, the added line of adaptivity and associated marks give little insight into the specifics of how the service will work as it changes over time. In fact, it even confuses the horizontal time scale in the diagram by placing a scale of many service encounters in between two scales of single service encounters.

To improve on this design, I suggest the creation of a tool to allow designers to specify several service blueprints, each for a different phase of execution. The tool would then automatically animate the changes between the different phases. For example, if a step of the service is removed over time (streamlining) then the block that represents it could fade out. The use of animation in this way would help guide the eye to notice important changes or other details that might be missed by a less dynamic design medium. Additionally, it would be much more effective than the line of adaptivity at communicating changes over time since it uses a more natural mapping.

The system I propose is not without its drawbacks. It would take significant effort to build such a system in a way that maintians its easy manipulation by a designer and thus effectiveness as a design medium. Additionally, it might take longer to compose a blueprint, and dynamic viewing would preclude printing on paper or other static media.

It is possible that such a system exists already. Does anyone know of something like this?

chigusa said

at 11:56 am on May 21, 2010

As there are many kinds of "planning software" available, (sometimes very expensive though), I will not be surprised if there exists such ones. Historically speaking, the first generation of simulations containing many elements were done in "operational research" area in the 50s to 60s, which could be said as a precursor of the system Andrew is pointing out. Thus "planning something with multiple elements and conditions, including logistics and work flows" has a long history from the very beginning of digital computing in many real world application areas.

Rob Ryan said

at 7:21 pm on May 21, 2010

-- whoops, meant to post this earlier --

This week's paper, concerning J. Forlizzi's Snackbot project, is simply puzzling. I'm not sure how to interpret the context in which Forlizzi et. al would be producing a project summary of a snack-delivering robot. At first, I grew excited for my friends in robotics -- someone had solved indoor navigation! -- but it turned out to be only "semi-autonomous." Huh.

Clearly, this 'paper,' is operating in a different modality: it seems more like a milestone document for a grant project. If so, then I think what this paper taught me most is about the differing traditions of research between these two Universities. The comparatively ontology-heavy (relative to Stanford) language and enumeration (e.g. -- the "three types of cultural model") have a distinctly foreign feel. There seems to be something serious brewing between design and HCI researchers: this is the second CMU professor to assert a distinctly third-party role for design. This document contains an intensely ethnographic and humanistic observation. The models are free from many of the constraints imposed on the typical Stanford study -- what does this say about differing traditions of research?

Now that I've realized this document is an exemplar of another sort of endeavor, I hesitate to render a judgement on this effort. In trying to imagine the trajectory of this project, I'm really only comfortable critiquing the end result of such an effort. I'm fascinated to see this talk, then -- it should be a bit uncomfortable.

Steven Dow said

at 11:55 pm on May 23, 2010

@Andrew: I like your ideas for adaptive service design software and I encourage you to develop them, potentially as your final report. Imagine some scenarios for how services may change and "animate" over time and describe the roles for software designers, service receivers, and service providers. What's the backbone of such a system? What happens when interaction breaks down?

Steven Dow said

at 11:56 pm on May 23, 2010

@Rob: A blog post like that deserves a Part 2. What were your post-class thoughts?

Nina Khosla said

at 2:35 pm on May 24, 2010

@Rob I found your comment "intensely ethnographic and humanistic observation" interesting, mainly because I didn't agree. I felt in terms of "ethnography," the insights were simplistic, something devoid of research altogether: "People want to eat snacks in their office," and "Services change over time." Somehow, I found myself not very inspired by the project itself. There was an edge in the paper, as if there's a "right" or a "wrong," a very specific way of doing things that I found frustrating.

Nina Khosla said

at 2:38 pm on May 24, 2010

It looks like I forgot to add this so here are my thoughts (in two parts b/c I'm too wordy, I guess)
The thing about this paper that most interested me was this idea of orientation and incorporation. I'm very interested in myth and narrative, both from a cultural and individual perspective. I posit that, in fact, you can create real change within people simply by trying to modify what their internal narrative is about why it is they are doing something. Orientation, to me, speaks to this idea of figuring out why you are doing something. Incorporation is the development and retelling of this narrative, sometimes to the point of becoming almost mythological - possibly this is when emotional ties to a product occur. Since I've already spent some time picking on Facebook this week, I'll use that as an example: Orientation is this initial interaction with the service: adding your friends, seeing how they interact with mutual friends, hearing these cultural myths about Facebook is supposed to be used. Incorporation is this increased interaction with Facebook - here's where you figure out what it's useful for, and start to really "depend" on it. For example, using Facebook as a phonebook (http://www.facebook.com/friends/?view=phonebook) or leaving messages on friend's pages. But also, as you begin to use it, a real emotional connection to the function it provides - being close to your friends and being able to speak about your life to them - thinking about how people have personally, emotionally integrated this understanding explains much of the backlash against Facebook.

Nina Khosla said

at 2:38 pm on May 24, 2010

I was also intrigued by the idea of figuring out how to graphically represent this ongoing interaction with a product or service. I think one of the skills that is crucial for designers, is to try to understand situations by understanding it from many different perspectives. I think having some different diagrams and graphical representatives, and creating a basic sense within designers for how to represent information in different ways is important. One of the things that I wish wasn't implied in the paper was that there was one "right" way to present this information, but rather that this is one thing to consider in our diagrams of this information. Even within this, there are probably many different ways we can map this change and understand it.

Rob Ryan said

at 9:37 pm on May 25, 2010

@Steven --

Funny you should say that -- I felt as though I should re-post after class, considering the disparities I found between seminar and lecture. Clearly, Dr. Forlizzi was possessed of a different vision of design, one much more at home in my heart. It's a vision that privileges visual design over [mechanical/statistical] engineering, which has the edge at Stanford. She is also very invested in her vision, and offers a self-confidence about her research that is distinctly foreign to Stanford. So much of her seminar discussion (clarifications, answers, questions-to-the-students) were couched in aggressively assured language, free of academic qualification.

I found it refreshing. I looked up her program, and am seriously considering applying.

I also found the body language of all parties during the discussion of high interest. So much of Terry's non-verbal signals were broadcasting: "I disapprove of the way you view things." In all, it felt rather tense when differing fundamental beliefs surfaced concerning design: in a very literal sense, she comes from a different "school" of design.

@Nina --

You and I may share different notions of ethnographic research. I'm referring to what Clifford Geertz called "thick description" -- a bottom-up annotative view of people's lives that doesn't reach for top-down analysis. Forlizzi is fundamentally unconcerned with justifying her design choices: rather, she's interested in making objects that enhance interaction. I would guess that she would take offense at the assertion that she doesn't do "real research."

That is the same sort of "right" / "wrong" you, in turn, seem to find frustrating.

If you didn't see any high-level insights, consider Forlizzi's insights about system process diagrams, which are not taught explicitly at Stanford. Just because it's a narrative, I would argue, doesn't mean it's not instructive.

Nina Khosla said

at 2:44 am on May 28, 2010

@Rob Oh, I very much agree with that idea of bottom-up, there's no other way to do ethnographic research. The reason I find myself uninspired by her research was again, that even working from the bottom up, she just didn't seem to go very far. I should be fair and say that there were some interesting bits of research she did that she talked about, but on the whole, I thought the research just wasn't that engaging. It seemed to be simplified to the point where she could put something in an academic paper or speak about it to engineers, as opposed to dealing with a messy, richer datasets that go beyond observations about behavior to implications about motivations, feelings, and values.
If I were to do something in her vein of research, I would start with observing analogous situations - tools like Opentable and Fandago that allow for automatic, non-human interaction where human interaction previously existed - and then to physical versions of these tools - ATMs being an obvious starting point. I'd want to understand the convenience of these tools and why or why not people preferred them. I think this would give me some beginning intuition about the way people thought about interacting with robots.

Nina Khosla said

at 2:53 am on May 28, 2010

@Rob - I should be clearer here, and say why exactly I feel this would lead to richer insight. It sounds to me like she is observing on a limited data set, looking at specifically, how people react to her robot that she's created and put into the world. I view this as sort of, putting a random point on the map. You can move from there to other prototypes and ideas. But if you do a richer level of research, you can put a "better" starting point down. Both points can theoretically lead to the same end point, but you also need to rely on deeper human insight to get to a more nuanced product/service. By starting with a more thoughtful product, your framing of the solution predisposes you to deeper thinking.

Again, as to the systems process diagrams, I only found them mildly interesting, and while she did say something interesting things about them, I wonder how much of that was a side effect of using big words to describe the phenomena she witnessed to make them sound more interesting.

It's most likely I'm predisposed to being nitpicky about her work because of the way it's presented and the channels she's developed her research through. That push to publish in academic journals reduces the degrees of freedom in her research, and that's hard because so much of design research is based on intuition. Again, I come back to thinking about a method of displaying work that's more appropriate to designers.

Poornimaw said

at 7:55 pm on Jun 9, 2010

I was intrigued by her explanation of the "sense-making process" when people are adopting a new device / product into their lives. I agree with Andrews concerns that the Adaptive service design blueprint proposed in the paper is not really clear about how it captures or understands this sense-making process, when most of the changes in user interaction with the system would occur.

You don't have permission to comment on this page.