fields

PuBliC: Future Everything one day living labs (x2)

During the Future Everything festival February 2015 Design Informatics were invited to develop and deliver a 48 hour workshop for the festivals Future Lab strand. Our proposal entitled PuBliC originally intended to offer a platforms across which Pedestrians, Bikes and Cars (or other road transport) could cooperate toward better travel experiences. In the end we concentrated on Bikes and developed daylong workshops that explored the barriers to a better cycling experience. The overarching concept for the workshop was the development of a cooperative investigation into the frictions of cycling in Manchester involving three phases:

1. Participants would be loaned bicycles to explore the city, whilst also gathering qualitative and quantitative data through the use of smart phones attached to handle bars.
2. A paper would be written throughout each day and published at the panel authored by participants, members of the academic team and delegates to the conference. The paper would featured material gathered automatically from the smart phones and a questionnaire filled in by participants.
3. Reflections, findings and thoughts would be returned to a panel to be held at the end of each day at the Future Everything conference venue involving the research team, participants and invited guests.

Developing a framework in which the gathering of data, analysis and publication could be bound and involve participants and the research team within close connection was the aim. The methodological routines of design research have turned into a formula in which an ethnographic study is a chance to gather a few choice quotes to explain how a prototype fulfills the intentions of the designer. As Ingold says:

“Disguised as social scientists we enter this world either by stealth, feigning invisibility, or under false pretences by claiming we have come to learn from teachers whose words are heeded not for the guidance they have to offer but as evidence of how they think, of their beliefs or attitudes. Then, as soon as we have filled our bags, we cut and run. This, in my estimation, is fundamentally unethical. It is to turn our backs upon the world in which we live and to which we owe our formation. With all the data at our fingertips, we think we know what can be known: yet knowing all, we fail to see or take our counsel from the world itself.” Tim Ingold Making: Anthropology, Archaeology, Art and Architecture. 2013 p.5

Part living lab and part critique of participatory research, the two one day workshops were a chance to explore how the dissemination of data could be kept as close to the people who produced it as possible.

Phase one: Cycle probes

A suite of bikes was provided by Pavol Gajdos who founded the Manchester Bike Hire (http://www.manchesterbikehire.co.uk) and were placed ready for hire in Albert Square outside the Manchester Town Hall, where the Future Everything conference was taking place (Fig 1.). Participants could book a bike through an event brite listing or turn up at our desk in the Town Hall (Fig 2.). On arrival participants were asked to filled in three forms –1. ethical consent to data capture, 2. brief them on safety and 3. also ask them to complete a preliminary questionnaire that would provide basic data that would go immediately into the paper. By providing figures and comments to questions ranging from: “How much do you cycle?”, to more complex questions about the capture and storage of data by the team:“If you feel ownership over that data, what would you be willing to give away?” the information formed the statistical basis of the overview to the Participants part of the paper. As new members joined the workshops, data was aggregated dynamically and written into the paper.

Having completed the forms, participants were provided with a cycle helmet and high visibility vest as well as an iPhone 4 preinstalled with Comob Net software for capturing GPS data. The Comob Net app had been modified slightly from the one available in the App Store to allow for messages of up to 200 characters. Before heading off into the city, participants were asked to use the app to record their experiences. The team used the vocabulary of ‘friction’ as the subject for participants to report on. Friction represented the impediments or resistance to moving through the city on a bike and cyclists were asked to enter messages into the iPhone App whenever they encountered something, someone or some circumstance that made them feel nervous, unsafe, angry, frustrated or on a more positive note excited, engaged or amused.

As participants moved into the city the iPhones pushed their live location to the Comob server, which was then visualised on a large LCD screen in the City Hall. At the same time the server began listing the messages that people were posting as they encountered ‘friction’ in the streets. Both of these data sets began to provide material for the co-authored paper that was already underway at the start of each day. Samples of what people posted across the two days through Comob:

“I feel connected with Manchester. but I can’t feel my hands”
“exchanged raffle tickets for bike supervision”
“cars waiting in space reserved for bikes at traffic lights”
“cycling one way system the wrong way!”
“bollards, canals and geese…cycling is dangerous work”
“resistance is impossible without fuel”
“lots of places to lock the bike, well done Manchester”
“just had a considerate bus driver”
“withington route ready to go and off we go missing some Tarmac!”
“v hard getting x town w M/link works”

Phase 2: Co-authorship and publication

As Ingold suggests, so often academics “enter this world…under false pretences…  Then, as soon as we have filled our bags, we cut and run.” The template format of successful SIGCHI papers is beginning to suggest such practices. The uber conference that is now the zenith of even a designer’s academic career is turning participatory and situated research into a recipe for methods for data capture over design. In an explicit effort to close the gap between participant voice and academic reflection, the team developed an open web-based Latex platform that allowed anybody to both see and contribute to the production of a paper on each day of the workshop. Using the SIGCHI Latex formatting as the style guide, the paper intended to parody the best of CHI papers but support an engagement with workshop participants and fellow conference goers right up until the point of publication. The platform was developed by Hadi Mehrpouya and Karl Monsen with Debbie Maxwell and I designing sections and participation. The system primarily uses Etherpad (etherpad.org), which was designed for real time collaboration on documents. Etherpad in turn runs on NodeJS, which is a non-blocking Javascript interpreter branched from Google Chrome’s webkit engine. Karl and Hadi then wrote custom PHP scripts that pulled in the text from each ‘pad’ and autogenerated text from the survey responses. The script finally executes the ‘tex2pdf’ program, which is part of the Latex package. This compiles the latex template and the fetched text into a PDF file. The PDF is embedded on the page using the IFRAME tag and the HTML is served by the Apache webserver. This all runs on GNU/Linux.

Engagement with the platform / paper began at different stages of each day of each workshop for different people. Sections were predefined according to typical sections.

Background The design team including academics would clear the previous day’s writing and begin to develop a background section. We decided that all background literature that foregrounded the activities and their analysis, would only be derived from the talks and presentations that occurred on that day within Future Everything. This meant two/three team members sitting in panels and streams developing text that supported emerging themes that were present within the Latex paper. Often in different rooms whist this took place, there was no time to discuss themes or particular narratives – and only the opportunity to mediate and develop these throughout the paper writing process. On the first day the team spent a considerable amount of time describing the methods and background to the actual workshop, our intentions and tactics so material from the conference was limited. By the second day the team chose to repurpose these aspects to allow more of us to attend the conference and draw out more interesting theoretical material and related projects. At various points in the day social media was also used to encourage other conference delegates to contribute to the paper. A good example was Tom de Grunwald’s contribution that occurred through conversation and an email submission that was sent to myself following his interest in our process: “The paper, both in content and form, participates in the emerging area of co-owned, permission based data sharing, as exemplfied by Open Paths [8], an initiative of the The New York Times Research & Development group [9].”  Tom de Grunwald, Co-Author, paper 1.

8. OpenPaths https://openpaths.cc/ [Accessed 26 February 2015.]

9. nytlabs http://nytlabs.com/projects/openpaths.html [Accessed 26 February 2015.]

Methodology, Participants, Data Collection and Analysis Following a rationale for the methods that we were using for the workshop the paper pulled material from the questionnaire that participants completed at the beginning of their involvement with the workshop. Serverside calculations were made that fed directly into the paper: “All participants were familiar with technology, owned digital cameras, and 86.6667% cycle at least once a week.” “15.3846% cycle mainly for leisure, 0.0000% cycle for mainly sport, whilst 7.6923% cycle for short trips (e.g. shopping) purposes.” Of course the sums need some work, the aggregation of commuters led to over 100%: “Out of the participants who cycled, 114.2857% cycle mainly for commuting purposes”. Despite these glitches, the opportunity to integrate quantitative data into a paper which was updating all of the time, as people continued to complete the form was very exciting, and allowed the numbers to gain a level of performativity which inflected the balance of the writing.

Findings Participants were encouraged to write into the three sections that made up the Findings. Sub headings of these sections were decided as the days went along and according to themes that were emerging and participants chose which one to edit depending upon these themes and the comments that other authors were writing into each section. Interesting examples of editing can be found in paper one that involved the academic speculating upon messages sent from the Comob application that were then extended by the participant: Academic: Blue Participant: Red

Participants X commented how water resists resistance perhaps as though the presence of water on the road surface impeded braking, or something more about how rain affects an ability to take control of the road and resist the hegemony of the car. X stopped by the river, watched it beat waves towards the riverbanks, how it couldn’t resist the pull of downstream.

The participants extension suggests a misreading by the academic and instead offers a more poetic, less descriptive perspective. Participant W was much bolder in qualifying what was meant by his message:

Other materials on the road were highlighted by participant W who simply wrote ”horse shit!” and later explained the dangers of animal excrement that is often found on road surfaces. Participant W: “Autocomplete put the words in my mouth. What I actually screamed was ”Holy crap!””. Things in the way, physically and unseen, the pollution filling my lungs.

In paper two, 3 participants spent a considerable amount of time at laptops at the workshop HQ describing their experiences during the afternoon. Participant A made suggestions for the use of technology to highlight difficult roads:

The biggest problems for me are having very narrow roads where parked cars are on the left of the road with cars on the right going past very fast; and also junctions with multiple lanes of traffic and having to change lanes. If the data could show which routes avoided these roads / junctions it would go towards helping show what the most safe routes are. If the cyclist had sensors in front, behind, left and right, it would be interesting to see how much space cyclists have around them on different routes.

Participant C developed a reflection upon existing data capture from Google as well as an opportunity for gathering data more relevant for cyclists, whilst on the back of a tandem:

 The notion of crowd sourced cycle and transport data emerged whilst I was being steered around the city by my tandem companion. Google’s turn by turn navigation system uses massively crowd sourced data coming from its Android operating system to provide live information about traffic flow rates, this allows car drivers to circumnavigate traffic black spots. Similarly Google’s Streetview project not only allows users to view any included location photogrpahically, but Google’s back-end systems analyse the images in order to populate their address databases and other services. Their Re:Capture system is used to verify information such as door numbers and building name signs. Similar systems could certainly be applied to cycles: collecting data from cycle-mounted cameras and GPS devices and using that data to add value to other services. For instance the evasive action taken to avoid potholes could be shared, live, with cycles approaching that particular pothole. Potentially drivers of motor vehicles could be warned of approaching cyclists through cycle-to-car data links.

Participant E delved into the social interactions that they experienced in the city and in particular the frictions felt in the relationships with other roads users :

When we collided with a pedestrian, we cycled back around to apologise. He waved his hand and gave a wry smile. Setting off when the lights go green is often nerve wracking, will the vehicles behind give you space? If turning right then being stuck in the middle of the road, with moving vehicles on either side, waiting for a gap in the traffic is intimidating. Motor vehicles frequently do not respect the ‘advance stop’ boxes. I wonder whether they know the reason why I deliberately manouvure my cycle around their vehicle, stop in front of them, and turn around to glance at them?

Discussion and implications The discussion and implications part of each paper drew together the themes within the paper that had emerged both from the conference discussions and the findings gathered from participants and academics reflections. It also allowed space for reflection on the co-authoring methods that led to the paper, and used the closing panel to summarise both participant experiences and invite guests to discuss implications.

Phase 3: Panel

At the end of each day between 4.30 and 5.30 the team, participants who volunteered, conference delegates who expressed an interest and special guests gathered to discuss the day using each paper as a guide. Guests for day one included: Cllr Mandie Shilton Godwin, Manchester City Council; Rafael Cuesta, Transport for Greater Manchester, and Pete Abel, Manchester Friends of the Earth. Day two included: Steve Turner, Manchester City Council; Matthew Fox, Future Cities Catapult, and Rose Barraclough, Arts Council England. The composition of each panel and its guests offered an opportunity to pick over the days activities and experiences. 45 mins was just about enough time to explore some personal experiences and then for guests to extend discussions about the implications.

Media platforms complemented the debate, on day one we used the Comob visualisations to describe the extent of participants roaming, on day 2 video gathered from a participants wearing of Google Glass was used to complement the discussion. The Glass video was particularly interesting because it offered a particularly grounded perspective as panellists watched cyclists coping with erratic pedestrians, oversized lorries as well as passing through red lights in order to find safer places to stop on roads.

During each Panel Debbie worked hard to edit material as well as feature aspects of the panel discussion: Commenting on the co-authoring process Debbie added: “This framework provides potential for a more democratic and visible approach for research participants in comparison to completing surveys and taking part in interviews without ever seeing the results.” She also drew attention to an aspect of the panel discussion that critiqued the use of an academic paper as output: “Discussion however noted that conventional publication outputs may not be the most appropriate or creative dissemination mechanism.” Making sure everyone who wanted to speak had spoken the team declared that the paper was ready for publishing and the PDF within the Latex platform was downloaded and sent to the printer signalling the end of the workshop for that day.

Reflections

Given the complexity of the platform, the bikes and the mobile software the workshop proved to be a useful and exciting way of exploring a participatory model for a living lab – albeit in one day. The co-authoring dimension and publication of the paper at the end of the day provided a bounded model in which the research could be articulated quickly but in a meaningful manner.

The paper from day 2 was submitted as a full paper to the BCS/HCI 2015 conference. Topped and tailed slightly to explain its generation, the paper featured all 31 authors and we all currently await a decision from the review panel. Expectations are realistically low given the live editorial process. However the paper itself is intact from the workshop with only typos amended. Given the template HCI papers that seem to be on the increase we hope that the model disrupts the status quo that is drying up design, the representation of participants and the sense of context that should be present in participatory research.

The team of 9 consisted of: Fionn Tynan-O’Mahony, Deborah Maxwell, Jane Macdonald, Hadi Mehrpouya, Sha Li, Anais Moisy, Mark Kobine, Karl Monsen and myself. Gabriele Schliwa coordinated things from Manchester and provided vital contacts through local cycle organisations, invited speakers for the panel and with Pavol Gajdos who founded http://www.manchesterbikehire.co.uk and provided us with 15 bikes for hire with high vis vests and helmets.

Links to the two papers are here:

Day 1: https://chrisspeed.net/wp-content/uploads/2015/02/PuBliCDay1.pdf
Day2: https://chrisspeed.net/hadi/fe/day2.pdf