Conventional Laboratory Design Does Not Compute

To Support Computational Research, It’s Time to Start Sharing Traditional Resources

June 16, 2015

Principal / Architect, NBBJ

When people think of laboratories, they typically imagine the traditional “wet” lab: high benches arranged in orderly rows, stocked with beakers and test tubes, with sinks and heavy equipment within arm’s reach. But now, as big data and computation change the nature of discovery and processes for research, it’s time to holistically rethink the configuration of the lab environment.

Until recently, computation in research was purely analytic, primarily used to understand and make sense of physical phenomena, and to evaluate measured effects against hypotheses. Now, however, computation is both analytical and generative: experimentation that once required physical testing, at a traditional lab bench, can now be done in the computer — it’s the same work, only modeled virtually instead of physically.

Following this shift, many researchers find themselves in environments that don’t support virtual experimentation. I recently visited a new, highly publicized research building where, only a year and a half after completion, researchers had converted their “traditional” wet bench tops into computer workstations. Almost from day one, found themselves having to adapt to an environment that hadn’t anticipated their needs.


A traditional, “wet” bench laboratory. (© Frank Oudeman)

Today, the common assumption is that around 50 percent of the laboratory environment should consist of traditional “wet” benches. But when we visit new facilities, we see such low utilization that it probably needs to be only 20 or 30 percent. Researchers just aren’t working the way they used to, and they shouldn’t have to continually adapt environments to support their important work.

So maybe it’s time we think of the traditional lab, the “wet” bench, not as assigned space but as a shared resource. Perhaps it should be located in a separate space, shared by teams, with the majority of the environment designed for collaborative, computation-based research.

Researchers already share resources: walk any lab floor, and you’ll see neighborhoods of benches with shared equipment in between. Not everyone needs, say, a nuclear magnetic resonance spectrometer. We should think about the lab benches themselves the same way: not everyone needs constant access to 20 linear feet of wet bench for their work, not when they can model it in the computer.

The tech industry has had great success with this arrangement. The workplaces of companies like Facebook consist of large open spaces where people and teams work together, with shared resources like conference rooms situated nearby. In a research environment, one could imagine some of those spaces being “wet” labs, booked when needed. Not that laboratories should suddenly look like Google, but some features could be adapted and repurposed to create truly innovative research environments.

So let’s plan on building fewer “wet” labs and focusing more of our attention on the computational lab: what it looks like, how it’s arranged, what furniture to include, what infrastructure is necessary. It wouldn’t need overhead carriers or high benches, or the standard 10′-6″ modules, but should it be more like an open office or a private office environment? Or a hybrid of the two? Should it be more like a studio, a loft, or a hackathon space?

Regardless, the notion that the laboratory environment consists primarily of “wet” benches is no longer a given. The nature of science has changed. We should start building accordingly.

Image courtesy of University of Michigan SNRE/Flickr.

Share this:  envelope facebook twitter googleplus tumblr linkedin
Follow nbbX