Public Lab’s Keys to Developing Low-Cost Science Tools

This piece was co-written by Mathew Lippincott.

The Public Laboratory for Open Technology and Science community is a massive petri dish for low-cost science tools. Our balloon-mapping tool is in its mature phase having evolved out of the agar during the 2010 Deepwater Horizon oil spill. This success was due in large part to the feedback provided by the community of tool users and consumers of tool data and their revisions to the tool. As we’ve broadened our development, we’ve asked, how can this success be replicated with other low-cost science tools still in the petri dish?

Rather than looking at specific breakthroughs in the balloon success story, we must examine why the agar in the petri dish was perfect for cultivating such a tool. What environmental conditions existed that pushed rapid development? Four clear features come to mind.

Key to Rapid Development

1. An urgent need for the tool emerged: The public was hungry for images, and the government had implemented a restriction on flights below 3,000 feet, preventing image acquisition.
2. The right tool developers were talking to users in the field: We had first-rate programming assistance and an online community of map lovers to tackle problems as they arose. As an example, the need for easier image sorting, as articulated by users in the field, led to Mapmill.org.
3. Advanced consumer technology was easy to leverage: The low-tech balloon, wedded to cheap, ubiquitous digital cameras produced excellent high-resolution imagery, and the balloon is free of burdensome use restrictions when compared with drones or manned aircraft.
4. The data is accessible to a broad audience: Real-space photography is interpretable without substantial scientific training.

The urgent need for the tool is the most important factor — we aim to answer questions in communities that have environmental health threats. Without a use case, a tool is not a tool, it’s merely a tech demo— like a web-enabled toaster. The tool must produce data that answers a compelling question. In order to answer a community’s question, the answer, and therefore the form of the data, must share a few features.

how to answer a community’s question

Clear data by design: Ideally, no one should have to ask, “What does this mean?” Realistically, asking the question, “What does this mean?” shouldn’t be intimidating. Where possible, data should contextualized in real space — either captured as visual imagery or projected within real space during data acquisition. New information is more legible when couched amid a familiar situation; capturing and preserving the context of data collection should be an integral aspect of the tool and its workflow. Acronyms, jargon, and disassociated quantitative data should be minimized.

Allow users to be tool designers: Those who need data should collect it, and those who use the collection tools should be able to design them. Open source is central, but accessibility of the design process plays out in a variety of material choices — in our balloon kit, we prefer a rubber band over the camera’s shutter to re-programming the camera. Rubber bands aren’t intimidating, and don’t have an off-putting aura of expertise around them. Everyone suggests newer, better rubber band techniques.

Build data authority through engaging stakeholders: Authority is granted by people, not by any specific technical process. Data standards should conform to the necessary end use — if those who need to handle data accept it, it’s authoritative. If needed in court, then follow appropriate chain-of-custody requirements by engaging lawyers. If needed to compete with industrial data, engage industry professionals in how to meet their standards.

As our tool development process moves forward, the principles outlined here may change shape or be re-prioritized. But the core principle of progress with purpose will remain. A motivated community can make rapid progress, and if the environmental conditions are met, success is inevitable.

Update: This post stated the incorrect altitude for a government-implemented flight restriction. The correct number is below 3,000 feet.

Mathew Lippincott lives in Portland, Ore., where he works on design issues in sanitation through the Cloacina Project, is faculty at the Pacific Northwest College of Art, and designs civic science tools as a founding member of the Public Laboratory for Open Technology and Science.

Adam Griffith, a 2011 Knight News Challenge winner, is a co-founder of the Public Laboratory for Open Technology and Science and serves as their director of Science and Coastal Environments. He received a B.S. in Biology from Roanoke College in 1999 and was subsequently accepted to Teach for America. He taught 6th grade science in the Houston Independent School District in Texas for three years before becoming a kayak instructor and head raft guide for the Nantahala Outdoor Center. He received his M.S. in Biology in 2008 studying the native bamboo Arundinaria gigantea (rivercane) and continues to work with the Eastern Band of Cherokee Indians and community members in Western N.C. as coordinator of the Rivercane Restoration Project. Adam is currently a coastal research scientist in the Program for the Study of Developed Shorelines at Western Carolina University and is based in Asheville, N.C.