When Pulse Lab Jakarta (PLJ) first envisioned the concept of Lab on Wheels, there were two main goals in mind. The first, encoded into the name itself, was to promote PLJ’s data analytics work to a broad audience. Second, through a variety of activities engaging participants during these outbound junctures, was to advocate data literacy and trigger conversations on the use of non-conventional data in more diverse contexts. In other words, Lab on Wheels was a unique way to showcase how big data and artificial intelligence could be used to support development practices and humanitarian action through practical, hands-on sessions with diverse practitioner groups across Indonesia.
As the rate of Indonesia’s urbanisation increases rapidly, urban issues become more and more complex. This creates a challenge for urban planners to find effective solutions to address increasingly complex issues that seem to multiply rapidly as urban populations balloon. As homes to billions of people, housing centres of technological innovations and engines of economic growth, cities are also the home of abundant amounts of data. This offers huge opportunities for urban planners to explore unconventional ways of solving urban issues. That’s why conversations around the utilisation of big data have also become part of the emerging urban planning discourse.
Broadly defined within the umbrella term of big data, non conventional data (also known as non-traditional data or alternative data), includes data that are often difficult to be structured and come from a range of diverse sources such as satellite imagery, social media, mobile networks, sensors, etc. A closer look at urban settings may reveal useful resources that can help in understanding and addressing complex urban issues. This is how PLJ came to use Twitter data to infer Jakarta’s commuting behaviour.
PLJ’s latest iteration of Lab on Wheels happened during the International Society of City and Regional Planners (ISOCARP) forum, where they sought to introduce how big and non-conventional data could inform urban planning. This blog gives a first-person behind the scenes look at the process.
The challenges and concerns
We employed both an internal and external lens when figuring out how best to structure this specific event. Internally, we considered our audience. Our Lab on Wheels is generally intended to be a light, hands-on session to introduce ways to leverage non-conventional data sources according to the context. Realising that the audience would consist of young urban planners and professionals, we needed to tailor our content to their specific contexts and take advantage of the fact that their jobs already brought them into close contact with data on a regular basis. The academic spectrum of participants — ranging from undergraduate students to urban planners with PhDs — also needed to be taken into consideration when selecting appropriate segments of big data to be discussed. What aspects could be communicated to both undergraduates as well as PhD students which would excite both? Important also was setting expectations from the start, highlighting that that this would be a light (albeit hands-on) introduction.
The external lens came into play by ensuring the overall theme of the conference (“Beyond the Metropolis”) as well as the overall agenda of the conference (our slot was on the third and last day when we expected people to be tired with waning interest) also influence the design of the content to be delivered, in addition to the dissemination and audience engagement strategy.
We fancied the idea of having an interactive session built on and supported by two “studio cases” (one on Transit Oriented Development also known as TOD, and the other on the historical preservation of Jalan Suryakencana, Bogor) which were presented in previous sessions in the conference. We saw that ride-hailing data could inform the TOD studio case. We also decided to use the data “walkshop” and our own data toolkit as a problem identification tool. This time, however, we did not conduct the data walkshop as a distinct process, but instead incorporated the data walkshop as part of a process to begin thinking about potential data sources and predicting potential data collection points.
How do we begin to demystify “Big Data” for young professionals?
In devising a strategy to ensure we accomplish the goal of this Lab on Wheels, we took a look back at the key purpose of the session. We wanted the audience to know the uses of non-conventional data for urban planning as well as understand the limits. We needed to cater to the “what and why”, without going overboard on the “how” part within the limited 2.5 hours alloted for our session. By giving participants enough information to effectively answer the “what and why” questions, we wanted our session to be catalytic in triggering further efforts to research and answer the “how” question themselves.
Big data can be rather complicated to explain. With the limited time available, we were only able to provide a surface-level view of data, big data, and non-conventional data, introduced in a manner that was contextualised to the audience’s own domains and experiences. To do this, we chose a participatory approach where we picked one of our own projects and mapped it to the participants’ daily journey (also known as ‘user journey mapping’ in the parlance of our social systems methodology).
For each aspect of their daily journey, we pointed out possible non-conventional data sources. For instance, by ordering a ride-hailing service to the office, we highlighted potential data that could be generated by that activity and how when we aggregate the data exhaust of thousands of people going to work through these means. We are able to develop Origin-Destination (OD) high spatial and temporal resolution data. By ‘putting ourselves in their shoes’ we could contextualise our knowledge to their domains and the concerns that preoccupy their work. What do they already know? And what do they need to learn?
Answering these questions themselves helped participants to achieve a better appreciation of, and the importance of non-conventional data in their work. We also employed interactive storytelling which was more effective for the young professionals still commencing their careers. Furthermore, we varied the level of ‘seriousness’ of each segment, and also employed real-time polls, peer group reviews, drawings, visualisations, videos, and use of sticky notes and photos to increase audience engagement and uptake of knowledge. With a bit of creativity and strict time-management, a large complicated topic such as ‘big data’ was able to be disseminated in a simple manner.
Explaining data analytics to those without data expertise
This question occupied quite a bit of our thinking when we were preparing for this specific Lab on Wheels. We had multiple brainstorming sessions focussed on answering two inter-related questions: “How do we make participants understand the importance of harnessing non-conventional data for urban planning?” and “how do we add depth to the amount of knowledge transferred without losing their attention in the process?” The narrative that was built from these questions, became the basis of our flow.
We mapped the journey flow and identified five aspects that had to be catered to through our session. The first was the tone. While the theme of the conference was “Beyond Metropolis” much of the audience came from across Indonesia and not just from Jakarta. Hence it was important for the content to resonate with them and their daily regional experiences. Afterwards, we tried to identify specific issues that they would have to be addressed within the context of the two studio cases. Thirdly, we wanted to provide the data context i.e.the definition, the case study, and the data ethics behind it. Next, we needed to have an ideation phase where the problems that were identified could be tackled through the use of non-conventional data. Lastly, we wanted to a peer review session where small groups (we ended up grouping the participants into 6 groups during the session) could present to each other in pairs.
From this Lab on Wheels experience, these are some useful things to note
Iterate, Iterate, Iterate
The overall process required several iterations within our own PLJ team. This helped us to spot missing elements. We found that it was useful to conduct our testing with people from different backgrounds which gave us a better sense of how our approach would work in different scenarios. Our testing involved members from both our data science and social system teams. This internal round of testing provided us with feedback, ranging from the effectiveness of methods to the clarity of language and terms used to guide the workshop process.
Drawings can engage people
When we first came up with the idea of “sketching out the problem” in the workshop we were a bit concerned. From our experience, we know the drawing process can take some time, especially when there are many participants involved. Furthermore the drawing process itself could become redundant, if participants are able to clearly identify the problem statement from the start. After some consideration, we decided to give it a go with the optimism that the drawing process would help participants to vividly visualise the problem they wanted to solve. Alas, the approach worked out well this time around, and even helped to highlight group dynamics and strengthen teamwork among the participants.
A picture can speak a thousand words
When we first conducted an internal ‘data walkshop’ in our lab, we became inspired to further experiment with it in the context of “Lab on Wheels”. We asked ourselves, “how might we inspire the participants to find unconventional data sources for urban planning?” Our answer was to design something that could help immerse participants in urban settings and contexts that could help them identify potential data sources that could be leveraged. Obviously we could not build a live urban setting! But we managed to use photos to help the participants imagine the relevant immersive urban contexts that we wanted them to appreciate. While it was not an ideal substitute for an actual immersive urban setting, it was still sufficiently functional to help the participants to look closer at the urban surroundings and identify possible data sources.
Bring in relatable cases
After setting the context and identifying the problem, we did not jump right into brainstorming on the non-conventional data sources for urban planning. Learning from our testing sessions, we understood that it was important to provide some examples that could enable participants to imagine scenarios where they could utilise non-conventional datasets in urban planning and in what ways. To do so, our data engineer presented some relatable cases from PLJ’s portfolio, one of which was related to developing commuting statistics using Twitter data. Even if participants were previously familiar with big data, we noticed that presenting an actual, specific, real-world use case from our own portfolio helped to enhance the ideation stage for the participants. This further enhanced our credibility with the audience, as we weren’t just speaking “theory” but also demonstrating experiential evidence.
We were happy that our efforts in preparing and strategising paid dividends in how well the session was received by participants. Not only were we able to make this diverse group of participants appreciate how non-conventional data could be used in their own work, but importantly we saw how the process stimulated their own creative thinking to innovate their own work.
PLJ is grateful to ISOCARP and IAP Indonesia for trusting us with this opportunity. The participants’ enthusiasm, curiosity and creative ideas that came out of this Lab on Wheels session also inspired us to look further in utilising non-conventional data to help solve complex urban issues. If you are interested in knowing more about our past Lab on Wheels, check out this video that captures the activities from previous sessions.