Spitmap is the name that I ultimately gave to my end-of-year project at Camberwell College of Arts. The project was my contribution to the "PITCH" collaboration between Old Spitalfields Market and the college which occurred on June 11th-13th of this year. I experienced the project as a series of successive failures and frustrations, so documenting it here may help me to take some lessons away from the project.
The Brief
The aim was to reconnect Old Spitalfields Market (OSM) with its local community and with its history. OSM is a market in the historic Spitalfields area of London (right by Brick Lane) which is an enclosed market that is managed as a set of market stalls consisting of food vendors and arts, crafts, and antiques vendors. A section of the market was given over to the college to be used for its own interpretation of what the market could be or become. This was to be demonstrated through objects, publications, interventions and workshops that would examine Spitalfields through a lens of either narrative, environment, exchange or interaction.
Students then had to gather research on the area and its history and individually or in groups propose a project falling under one of the elective disciplines (the aforementioned narrative, environment, exchange or interaction) to address the needs of the brief. That sounds rather vague, but it could manifest simply as a publication (say a zine or a comic or other illustrated book) that drew upon the area and was sold in the market.
 |
| Looking down on some stalls in Old Spitalfields Market, London |
Research
My research for the project consisted of visits to the market and surrounding area, producing some sketches and taking lots of photographs. My frustrations with producing sketches for research would prompt me to undertake a separate location sketching project so as to develop a technique that would be both economical and useful to me.
Through my visits to the area and sketches and photographs I made as research, I realized that my primary interest was in the environment, particularly how people related to the space and the relation between the space of the market and the larger Spitalfields area itself. However, turning this into a concrete and feasible idea proved rather difficult.
I went through various ideas: from changing the arrangement of the stalls to give an organic, meandering market feel intended to evoke a more medieval space to notions of containing the larger space of Spitalfields within the market.
While researching the latter idea I came across the work of Armelle Caron who makes prints by taking aerial maps of cities, translating them into a series of discrete shapes and then rearranging those shapes, placing them into organized rows. In representing space in this highly linear fashion, Caron challenges the integrity of mapping. I experimented very briefly in this direction, looking at the map of Spitalfields in the same way.
 |
| Spitalfields map disassembly experimentation |
I needed some connection to the history of Spitalfields and I was looking for a way to situate the individual within the environment. I came to feel that the Spitalfields Roman Woman - an archaeological find of a woman's remains in a lead coffin dating to Roman times in the area - was a useful nexus point for the historical and the personal and I began researching her as well.
I further explored works by Cornelia Parker, particularly 30 Pieces of Silver and Cold Dark Matter wherein she had a shed exploded by the British Army and then suspended the fragments around themselves in an installation, lit by a single light source in the interior. I also looked at the work of Christian Boltansky and Alexander Calder. Perhaps the key takeaway from these was the ability to communicate multiplicity through light and shadow.
While I was excited at the idea of a sort of exploded three-dimensional map installation that would be illuminated, I was also daunted by the prospect of such an undertaking and I had difficulty tying down its aims in concrete terms. Further, as I was encouraged by my tutors to focus on the feeling desired from the audience, I ended up leaning toward something more experiential, since it is difficult for me to understand feeling of anything in relation to sculpture or installation (certainly not in terms of an intended feeling).
Initial Proposal
My initial proposal was tentatively titled Map and Tomb and would consist of a replica coffin to the one the Roman Woman was found in. People in the market would be encouraged to step into the coffin, which would serve as a kind of sensory deprivation chamber, isolating them from the surrounding market and perhaps giving a sense of connection to the history of the place after contemplation. There was also to be an augmented-reality (AR) portion to the experience, which was not well-defined, where people in the market could explore a sort of alternate market through an AR app for their smartphone, with the app eventually leading them to the replica coffin.
In preparing my proposal I undertook making a scale model of the coffin that I was going to build. Since this would be completely new for me, making a scale model first seemed a sensible approach. I also needed to make some headway on the project, and that seemed to be the lowest hanging fruit, as my research into AR app development yielded far too many approaches and tools for me to quickly sort and no clear and simple paths forward for someone essentially foreign to software development.
I made the model using what I had on hand: some hard paper sheets that I had received along with some free paper samples. I determined the size from observations of the coffin in the Museum of London and the given information on the height of the woman. I then replaced the woman's height with that of a 95-percentile male (SAE) and scaled the coffin up accordingly.
With the basic shape of the coffin made, I set about recreating the intricate ornamental detail found on the coffin. This would be the key way of distinguishing the coffin, as otherwise it contains no historical mark. I used my drawings from my visit to the Museum of London as a guide, building up the scallop shell design with layers of card paper that I then scored with a bone folder to give the impression of indentations.
With the model constructed I went about painting and weathering it. I simply used mixtures of three colours of acrylic paint: Titanium White, Process Cyan and Yellow Ochre along with some black ink.
Once I had delivered my proposal it was time to execute it, and that's where reality started to set in. I was on a strict budget, as only so much money was allocated for the projects, so I needed to be resourceful with material acquisition. While I was initially hopeful that I would be able to find cheap or free materials suiting my purpose or an inexpensive, plain coffin needing only detailing, I was simply unable to come up with any solution for next to no money. I needed rather large continuous sheets of MDF or wood or some suitably stiff material, and things just weren't adding up.
Re-envisioning the Project
I felt like I had invested a lot in my conception of the project and got stuck when I was unable to make headway. I was able to make progress when one of the tutors, upon looking through my sketchbook, remarked that it seemed I worked by building up lots of little things rather than doing anything large and singular. This resonated with me, and I recognized the potential to simply realize some of the contents of my sketchbook in the market. I could do this by making scale models of historical objects from Spitalfields and setting them up around the market. Because these objects alone would call for an explanation, I would bring back the idea of AR in a more modest form, to serve as a way to bring up contextual information as one encountered an object. I had already built one object - the coffin of the Roman Woman - so I simply needed to build more (ideally many more) and work out the AR details.
For the other objects I started by turning to some research I had made but not incorporated - the Copperplate Map. The Copperplate Map is the earliest true map of London, dating to the 1550s, and part of Spitalfields can be seen in the Moorfields plate which is one of the few surviving plates. I liked the little windmills spied in the map and went about making them with the thought that I might get more such historical objects from the map. There was certainly a degree of appeal to me in excavating an object from a historical map (a 2D representation of space) and placing it within the real 3D space which had been represented as it now existed.
 |
| Moorfields plate of the Copperplate Map [Source] |
Model-Building
With model-building I came up against a number of problems. For one, I had failed to accumulate the research that would support my approach - I didn't have a vast collection of historically significant Spitalfields objects that I simply needed to physically realize. This was of course brought about by the freeform nature of the project, where research was conducted without a clear aim, and by my late game change of course. Another problem was that I actually had no experience in model-making. That the coffin went as well as it did was due to the sheer simplicity of the object and no doubt some luck. I certainly didn't have the time to invest in developing, or even researching for that matter, model-making techniques or best practices. The major problem was simply a lack of time - model-making can be very time-consuming and if I was ever going to execute the project satisfactorily I really would have needed to schedule the building of the models well in advance; I would have needed to know the desired outcome at the start of the project. Knowing these issues, I decided to dive in anyway, not wanting to wallow in defeat, and simply get as far as I could with the limited time I had remaining. I ordered some paper, bought some glue, and went to work.
Regretfully I didn't do a very good job of documenting the build process. For example I have no photos of the completed and painted Christ Church scale model (easily the most difficult model to make).
The Realities of Augmented Reality
With the model-building underway I needed to figure out the AR portion of the project. App development was off the table, as not only was I completely unacquainted with it and intimidated by the prospect, but the need to get certification from the App Store to avoid a cumbersome installation process for users and the time it would take meant that I simply would not have enough time. I looked at various AR apps that let you create and share AR content but found them to be intrusive and difficult to adapt for my purposes.
Fortunately I came across the work of Jerome Etienne, whose work on AR.js elegantly enables basic, fast AR functionality simply through a browser. Using AR.js, one can display simple objects, text, and animation on a phone with just a few lines of javascript. It was ideal for my purposes, as the use of a website would mean that people could simply direct their phones to a web address (no obnoxious downloads) and it required minimal coding (scripting really, if that) for me.
I knew that I needed a website for users. This would need to be a simple domain name to increase the likelihood that people would actually go to it, not a long string as offered by free domains. I decided against using a barcode as it seems that these are now considered dated, people seem to rarely engage with them, and on Android there is no built-in barcode recognition (you need a separate app for it whereas the iPhone built-in camera app recognizes barcodes fine), making the likelihood that Android users would engage with it effectively zero. So I went about purchasing a domain name - spitmap.uk - and I was off.
The simple example that I followed for AR.js uses an image called a marker. When this image is recognized in a smartphone's camera while it is on the website, the AR routine is initiated. All I needed to do was to make my own set of custom marker images. I decided to make a separate image for each object. That way I could display the images next to the object, users could then point their phones at the image and get contextual information displayed about the object.
However, in order to use my own custom marker files, I needed to be able to host them on a server so that I could point to them in code. I also needed to be able to put custom javascript code on my website. The domain that I purchased only allowed editing through a restrictive website builder application and did not offer file hosting functionality (at least not in a straightforward way where you can point to a file with a simple address). Most domain and website services are similar in this regard. I cast about looking for a way to host my files - Google Drive, iCloud, Dropbox - to no success. And of course I still needed a website that I could run my code on. All I had was a name, I was running out of time and the project seemed to be going up in smoke.
As it turned out a solution had been staring me in the face. The AR.js code that I had been so happy to poach for my own project was hosted on GitHub. On GitHub with a free account one can not only host code, but also files, since code projects typically need to access files. And the GitHub file hosting is a straightforward file-server. Better still, GitHub allows you to set up a website which has full functionality - you can run whatever code you like on it - since users are likely to use such sites to demonstrate their code. This is all possible with a free account. Perfect! The website you get with GitHub is a sub-domain of github.io, but since I had bought a domain name, I was able to redirect that site to my GitHub site (in the address bar the user would always see spitmap.uk rather than any github.io domain).
Once I had discovered GitHub as the answer to my problems I was able to set up a basic program to recognize my custom markers and then display an image and some text as the content. I was pretty much out of time at this point, and therefore wasn't able to put much effort into the content of the contextual pop-ups. While I could have conceivably mocked up more ambitious content in advance, I think leaving it until I got a basic frame working was the right call, since there wasn't much point in dreaming up possibilities which I had no idea on how to execute.
Difficulties on the Day
With so much crunch occurring in the final stretch, it should not surprise that the first day I encountered problems with setting up. Once I had set things up, I realized that some of the AR displays were flipped so that they were unreadable or difficult to read. I had to go back and edit the code to fix the problem. More significantly, I encountered a critical problem resulting from a crucial oversight on my part: the unreliability of the Wi-Fi network. It was known that there was Wi-Fi for OSM, which helped to justify my AR-website, but at least on the days of the event it was horribly flaky and unreliable, constantly disconnecting. The AR requires a fairly robust connection to work, and on the OSM Wi-Fi the website would behave in a buggy manner.
The AR also proved to be notoriously battery-hungry, which made troubleshooting it early on with my phone a little concerning.
Closing Thoughts
Perhaps my biggest regret on the project is my failure to capture good documentary evidence of the progress. This seems like a curious omission, as I did a much better job at documenting my earlier course project back in March of this year after making documentation a priority. In part this was due to a lack of a sense of direction I felt through much of the project. Without a clear objective, it was difficult to find a thread through everything; I frequently fail to document small experiments and trials when I don't have expectations for their success. It was also a consequence of too much work being back-loaded: I accomplished a lot in the final few days and I simply didn't have the time or presence of mind to record what I was doing. Had a been on a clear path from the start this would have been much less of an issue.
I think my initial takeaway from the project was the uselessness of purely exploratory research, of the need to set down constraints early so as to provide a direction for things. Time and again I was urged by the tutors not to get locked down or committed to a particular direction and to explore possibilities which I ultimately found futile. In short, I wanted to start the project with a clear sense of what the outcome should look like.
With some distance from the project, I think I've come away with a more nuanced and more valuable takeaway, which is about my own working method. For one I realized that I don't accomplish a task by planning it all out on paper and then simply executing it, since the execution part is what I am typically clueless about; rather I build up to things by making small things on the way, things which are either parts of a larger whole or iterations on the way to a more refined thing. It's not so much that I am adverse to exploration, but that exploration has to be deep, it has to go a significant part of the way toward the realization of the idea, for it to be very useful for me. Exploring in depth also means endeavoring to make each exploration worthwhile in its own right. With my early location sketches I found that they seemed a waste of time, since I didn't really use them nor could I see a way how to use them in my final outcomes. With my personally-assigned location sketch project, I sought simply to make each sketch count - in other words, just make a nice sketch - and I found that doing so enabled me to produce research that contained details that I would then later want to reference or incorporate. Perhaps the tutors simply take for granted that this is what students are always trying to do, hence why they push for producing quantity and working quickly, but I find for myself it takes a conscious choice to invest in research in a way that will later prove meaningful.
There was a certain pleasure in being able to build simple models, working out the plans on paper and then realizing them, although it was not without its frustrations, particularly as it relates to working with glue and waiting for it to dry. I should have liked to have worked out model-making to greater proficiency before starting down this path, preferably using a lighter weight paper so as to obtain better detail at the scale I was working and possibly using more appropriate adhesives. I wasn't overly pleased with the effect of the paint and I think a nice unpainted paper texture would look quite well and lend the objects a certain unity and also ethereal quality that would tie in to them being historical objects.
It's fairly typical for me to do a lot of reading as research, but I didn't do that for this project. In hindsight I think this was a mistake. With prior projects I found that reading took up a lot of time, giving me insufficient time to incorporate much from it into the work, and that what I did discover often seemed unusable, being something that I did not have the skill or time to execute in the time I had for the project. I purposely skipped reading for this project to give myself more space for other projects. However, doing so illustrated the value for me in incorporating reading into my research. It enables me to make connections that I would not otherwise make, provides me with ways of thinking about my work that if nothing else can help me to persist with an idea (something that could have been valuable on this project), and points me in useful directions. I realize that reading and writing is such an integral part of the way that I think that it is a necessary component of working through ideas, particularly when taking on unknown or difficult challenges.
Finally, I note that there were several course-run workshops throughout the project which involved experimentation with the research material collected. I did not find these workshops productive or helpful and have chosen not to include a discussion of them here.
Addendum
As it turns out, despite my having cancelled my domain service immediately after concluding the project, as of this writing going to the address spitmap.uk still redirects to my GitHub page. Going there on a device with a camera will activate that camera, and pointing that camera at one of the markers should bring up the AR content that was displayed in the market. So if you go to that address on your phone and point your phone at the markers shown on this page, it should work out.
