Thursday, June 14, 2018

Here is a cool viz

Here is a cool 3D viz made by a research group in Thailand

These are great, although I think the real work will be in how to use these to improve the campus. Seeing the data is one thing, using it is another.

Another paper using mobile phone data to understand travel behavior by the same crew. This work is similar to mine, although mine asks different questions.

Phithakkitnukoon, Santi, et al. "Understanding tourist behavior using large-scale mobile sensing approach: A case study of mobile phone users in Japan." Pervasive and Mobile Computing 18 (2015): 18-39.

Friday, May 4, 2018

Adults, seniors and children: Using maps in mixed group discussions in urban planning

Sidewalk Labs Controversy - Using Data for Design. Really??

Did you see this?

Its been all over the CBC this week.

I dont know a lot about the tracking of garbage, water and other utility systems, but I know a lot about tracking people in public spaces in order to design them better.

One big argument I hear on the radio is that these data will be invaluable to urban planners and architects across Canada.

I wish.

I am a believer in using data for design. Its what I do. Not by philosophical principle, but by NECESSITY. The data is coming, and by golly the designers better keep up!

I believe designers have got to step up, and become data experts. If you are a designer and you don't know what to do with data, get ready to be bulldozed by manufactures, technologists, and smartphone apps. Basically by those who stand to gain a profit, and who know how to use these data to do it. On the other hand - you COULD take a statistics course or two, and get with the program. :) Just sayin'.

There are a few points about this discussion I hear on the CBC that I think are being missed:

  • People seem to forget that we are already being tracked. Everywhere. We are being tracked by our phones, in public spaces, in retail stores. etc etc etc. I am really not sure its such big a deal to be adding it to this Toronto suburb.
  • In order to make powerful analysis we need a lot of data, and there needs to be some connection between these data - such as a user ID, in order for the data to be used for marketing or politics or other subversive manipulation. Urban data does not often include this. (although much of this data is being collected from our phones already - so its a moot point.) So while I think it needs to be regulated, its important to know particularly how the data can be linked to other data.
  • Data has been used to manipulate us in subversive ways, NOT by finding our personal information - but by finding MANY many peoples information, that are all linked through personal ID - and then looking for patterns in that. One person's data isn't really all that powerful for finding generalizable patterns - or subversively manipulating populations. Likewise, it isn't particularly helpful to know how people use ONE park. We need to know how multiple parks are used to see patterns.
  •  Actually for research purposes it is already ethically considered OK to collect data about people in public places as long as people are under the understanding that their behaviors can be seen (i.e. bathrooms are NOT OK. obvi), AND that the data being collected has no personal identifying features - i.e. faces. In other words the city or researchers could already be collecting this data. (They might be, but it is expensive to do and most designers don't know what to do with it once they've got it...on to my next point...)
  • The other big issue is that these data are being claimed to be helpful to planners - but generally, designers, are a little stumped about how to use these. We have been collecting similar data for decades, and are still struggling with how to use it to really benefit society. (I'm not saying it can't be used for society's benefit - just no one really wants to admit, that they don't really understand HOW - yet)
  • It is particularly useful to collect data from multiple similar places with the same design features with specific measurable benefits. Then we can see patterns. I am not sure the data collected in this one little suburb will be enough to make the kinds of claims they are making. Also, the questions being asked about human benefit are not easily tracked or answered, and we would actually need to figure out the questions BEFORE we start collecting, because we could be collecting ALL THE WRONG DATA! (Other PhD students - you know what I'm talking about!) 

Overall, research has made some conclusions in various published papers over the last decades, and made some headway in design but it is still a fairly periphery approach. Correlation doesn't mean causation, and it is difficult to measure real holistic benefits of urban planning such as public health or happiness. It is easier to use these data when money is involved - i.e. in retail - to find out what designs make people buy more - which is likely why it is getting funded and receiving so much attention.

BUT using data to create "better" urban designs? This, is not all that clear.

The main point is, planners and designers are not just sitting waiting for large data to answer all their design questions. Most have no idea what to do with this data, and the ones who do have already been  trying to use it, (with varied success) for decades.

Thursday, March 29, 2018

Is architecture an art or science? Taking hints from data viz expertise and HCI.

Personal reflections on architecture and human-centered design/data viz and data driven design.

Sheelagh Capendale, an expert in data visualization, whose work has been revolutionary in exploring how researchers investigate human behaviors in computer science, refers to her work as “observation”. She claims that she works in science, but she learned to observe in her background as an artist. She describes her experience as a young art student, in her first photography course where the instructor gave the excruciating assignment of asking students to photograph a blank wall; due in three hours. She refers to the process as a painful lesson in observation, that it wasn’t until after 2.5 hours of studying that wall, that the photos started to emerge. That after hours of looking, she finally began to “see”. Interestingly, she claims that she takes this same skill, the ability to “see”, into her work as a researcher. That this skill is as useful in her career as an internationally recognized scientist. Here, instead of photographing a wall, she observes human behavior. She observes how domain experts, and (at times) non-experts use visualizations of data. She uses multiple methodologies to assist her in her observation: quantitative qualitative methods, established and new. Even though this work is considered science, she uses the same skills developed as an artist. Here I argue that this act of “observation” is a learnable skill that and that adopting the posture of observation is admirable across disciplines, and in particular, I focus its application in the practice and theory of architecture.

Architecture is undergoing a paradigm shift. This shift is likely propelled by post-modernism, a data revolution, new modernism and a mass cultural shift towards pluralism, all of which is further bolstered by the ubiquity of social media and the communication age. This shift parallels the conceptual leap required of scientists in the centuries following Galileo’s declaration that the earth is not the center of the universe. The term “human-centered design” is not coincidental. For centuries, architects have not questioned the assumption that their perception of the universe (no matter how altruistic, or well-meaning) is the center of their inspiration. And here it is folks: it is not.

In my experience, the mainstream of architectural approaches do well at observing inanimate objects. Architecture theory focuses on materiality, the generation of form and expression the poetic as a spiritual marriage of idea and expression that will somehow inspire and elevate civilization through a universal experience of the architecture as expression. Louis Khan asked “what does the brick want to be?” Architects ponder the nature of materials as a contemplative, expressive, exploration of the emotive experience of texture, shape, and form. The power of form to express the poetic is a well explored topic in architecture schools.  I feel this has resulted in many architects who have an uncanny level of skill and ingenuity with materials and an ability to consciously manipulate the emotive, experiential effects of space, form and visual information. As researchers, this same approach is manifested in the philosophical approach of phenomenology. An approach that highlights and emphasizes the bias of the observer. The personal, self-reflective experience of space is the source of artistic inspiration, and the center (or starting place) of knowledge. This is where, as I see it, the trajectory of architectural theory has taken us. The architect at the center. A self-reflective expert, who then projects his inspiration on the world through his expression of experience through form.

While acknowledging bias is part of the process of becoming a great observer, it is not the end point. True observation is deeply concerned with understanding and learning the reality of what lies beyond our own bias. Confirmation bias is a knowledge killer. The reproducibility crisis in psychology research attests to the challenge of assuming we know without double (or even better, triple checking). The dangers of this can lead to mistakes. Mistakes made by people who are making decisions about cities, hospitals, medicines, health, life and death. It’s not good.

In data visualization there is a concept called “change blindness”. It is a well-researched phenomenon where people will often not see something if they don’t expect it. In studies where researchers replace actors without any hint of the replacement, or where there is a visual change in an image or scene, participants who are focused on something else, or not expecting the change - will not actually see it. The type of observation Sheelagh Carpendale is referring to is one which uses all available tools to overcome the natural tendency of change blindness. It is being a good observer, acknowledging bias, grappling with one’s own ego to get beyond what we think is in front of us - and ACTUALLY SEE WHAT IS THERE. It is a personal practice as well as adopting iterative, painstaking, rigorous, research methodologies to ensure our observations are as close a model of reality as we can hope for. Interestingly, while artists might be excellent at the former processes of self-reflection and emotive expression, scientists have well established methods for the latter. Although, as one moves further into the understanding observation, these divisions between disciplines is less useful. Artist, researcher, designer or scientist, the act of well-honed observation is critical to acquiring knowledge, understanding and innovation.

You may assume that true and great inspiration comes from the ethos of the artist, designer or great scientist. One may think this, because it is the narrative we have been told. The master (male) artist, the great (male) architect, the genius (male) scientist, who ponders existence, and finds inspiration through some magical moment of inspiration. But this narrative, is, of course not true. Inspiration and innovation is arrived at through a well-honed practice; that includes self-reflection, observation, craft, time, pain and struggle. This is inspired by a sincere and profound sense of curiosity. And though methods and objects may differ, this process is the same whether a person is a scientist, researcher, artist or designer.

All this said, architecture is a deeply inquisitive practice and architects themselves are altruistic and humble...or at least the ones I know are. Contrary to what I have just said, architectural practice also has a long history of using data and methods of inquiry in their practice, although this process has become periphery. While the discipline has focused on the generation of form, one might wonder where the “humans” are. Vitruvius in this seminal work about the philosophy or architecture talks about the three pillars of architecture: beauty, function and structure. While much of architecture research has highlighted the expressive nature of structure and beauty, the practice of human observation responds to the pillar of function. And while architecture research has yet to evolve into the established academic institution familiar to the sciences, the discipline has persisted by often disseminating new knowledge in neighboring disciplines such as urban planning, geography, philosophy and design theory. While this makes it difficult to define what architecture research is, it points to a large amount of relevant, data-driven research that has impacted architectural design culture and practice for generations... (this all leads to a discussion that I plan to continue to write about...stay tuned :) on using spatial data in design, the love/hate relationship with environmental psychology, the somewhat irrational fear of science on behalf of architects, and  awesome new relevant work that is generating new technologies and methods for understanding complex behavior patterns using hierarchical regression modeling and spatial data)

Architecture research needs to incorporate rigorous empirical methods using quantitative data. Sheelagh Carpendale is celebrated for her work in advocating for using multiple mixed-methods in data visualization. This is not an argument for which method is better. The nature of the inquiry and the research question dictates the best method to adopt. I argue that architects and architecture researchers should follow a similar trajectory as human computer interaction, and data visualization - while bringing a unique-to-architecture layer of design expertise. And while architectural research has primarily focused on qualitative processes of inquiry, it is time for architectural design researchers to balance out their practices and learn how to use (or at least benefit from) the tools of quantitative research. This balance of methodology will help tip the balance of architecture research towards the objectives of “human centered” design.

Friday, March 16, 2018

Architecture and human centred design:thoughts Post#1

It appears that in the architectural design process there are a number of existing processes that would suit various human-centered design practices. The mechanisms for these processes already exist and are included in the fee structure, and hence are prime opportunities for user centered activities. The phase that presents the most opportunities for user centered design is conceptual design, although there are opportunities later in the design process, particularly in design development where existing designs can be vetted, analyzed and evaluated by usersand the design team.

Typical human-centered design activities are performed in two key settings. One is among designers themselves who use their cognitive understanding of users to evaluate designs and form in-house criteria. The other setting is when designers work with stakeholders, clients and users to both evaluate designs and to develop new understandings about how users will interact with their buildings. There are multiple opportunities in the typical architecture process where this might occur. 

While it is easy for design team members to meet in-house, the stages where these types of activities can be performed include design reviews, design analysis, collecting background data, developing alternate designs, and the various design phases where the project is being developed.

Existing phases in the project where there are opportunities to engage with users, clients and stakeholders are more prevalent in the early design stages, but do continue throughout the process. A key consideration is post-occupancy studies, which may not be of interest to the client per-se but is crucial to the design team’s professional and the building of expertise within the firm. 

Concept Design 
Concept design presents the largest opportunity for human centered design. This can include gathering information about the users from second-hand sources such as media and historical documents. It can also include observational studies where designers observe users in their existing environment. This is the stage where developing a clear outline of the users and stakeholders is critical. Involving stakeholders in the process of outlining users, their network and relationship to the project and to each other should be done at this stage. Activities that help designers develop empathy for users of a particular demographic are important at this stage such as personality profiles and scenarios, outlining a day in the life, and network diagrams of relationships. While many of these can be done by the designers themselves it is important that some of these are practices with sample groups of users themselves to ensure assumptions are checked before designers move forward in the design. These can be integrated into design meetings, community meetings, interviews and questionnaires that are already present in the current fee structure.