Open Data & Wicked Environmental Problems - Building intention through Data Visualisation

Short Description: How do we look at environmental issues where data & sources are absent? What role can open-data, community building, & data visualisation play? This talk will share learnings from the India Sand Watch story, an environmental accountability project looking at the issue of illegal & destructive sand mining in India’s rivers.

Context: Solving for wicked environmental problems is not easy. There’s a severe lack of data, data sources themselves are absent, structures for implementation don’t exist, and a will to collaborate is either missing or has not been made possible for various reasons.

With Veditum’s project - India Sand Watch - when we picked the issue of illegal and destructive sand mining in India’s rivers, we started with the assumption that identifying illegality itself would be enough to solve the problem. We started with trying to use machine learning for this effort. Once we have the required data on mining activity, that should be enough to address the problem, right?

But we realised, thankfully early on, that this wouldn’t be enough. We had missed identifying one of the most crucial reasons that limit solutions to wicked environmental problems - intention. What we really needed to do, to solve a wicked problem like this, is build intention. We went back to the drawing board, to reimagine our response. And this response had to be designed with collaborators across non-governmental organizations, private citizens, researchers & academia, the judiciary, media persons, and state actors in mind.

The data visualization component: What followed? An open-data archive built on the Dots by Ooloi Knowledge Platform. This archive (with over 50000 data points now), accessible at sandwatch.in, is designed on the Sites of Violence (SoV) Framework in sand mining (developed by myself), that takes a systems approach and visualises the sand mining sector as a series of sites of violence and / or violation. A first presentation explaining the SoV framework can be accessed at this link - https://vimeo.com/586311487 (Starting around 13 minutes).

Once we had the data infrastructure in place, we went to work on building the data pipelines through scaled volunteering efforts, data sprints, partnerships, and more. Each of these efforts had a component of data visualisation - from zines to sprint report cards (which includes a playlist - ref link: veditum.org/datasprints) to GIS based maps, and more.

With the data, we’ve created visual evidence for successful court cases. Further, with partners from UC Berkeley, we have a successful machine learning model in place that identifies mining activity using open source remote sensing data. But how do you convert this into evidence? And take this to people who don’t understand ML? We created a physical artefact that recreates this story as a flipbook.

Participant take away: Through an engagement with a mix of physical & digital artefacts, a string of learnings on how to use open-data & data-viz for wicked environmental problems. The talk will be useful for those interested in working on environmental problems as well as those working on data visualisation to understand what it takes to build open datasets.