Autonomy in Synthetic Biology

A lot of work is going into laboratory automation and design tools, but how far is it actually getting and what are the real roadblocks? We examine these questions in a new article, "Levels of autonomy in synthetic biology," out today in Molecular Systems Biology.

I'm a big believer that a better toolkit is eventually going to radically transform the way that we engineer biological organisms. And I've certainly preached the gospel of data integration, design tools, reproducibility, etc. But somehow, two decades after the field began, we still find ourselves with lots of barriers to effective deployment of standards and automation to actually increase routine productivity. Way too much is still slow, manual, artisanal. So what's going wrong and what do we need to fix in order to get that transformation into a world of rapid, routine, and reliable engineering?

To understand this, we first developed a six-level framework for analyzing efficacy of automation, analogous to the one used for discussing autonomous vehicles. We don't necessarily need to go crazy-high in autonomy in order to get a lot of benefit. What I really want is at least some good Level 2 scientific "driver assistance" features to help me with the lab equivalents of lane-keeping, checking my blind spot, and parallel parking.

State of the art in autonomy for design-build-test-learn cycle, as shown in our article.

The problem is that right now, there's a bunch of good work being done on specific challenges in the prototypical "design", "build", "test", and "learn" stages of the engineering cycle, but not enough investment in the "glue" of standards that will allow things to connect together between stages or in curation tools that decrease the burden in setting up tools. We know from a number of demonstrations that we can do much better, but the marketplace is still too fragmented and just not enough work has been done on stringing these pieces together yet.

At the same time, I have a lot of hope.  The work we've been doing in partnership with lots of others on the DARPA SD2 program is producing a lot of interesting tools for lowering barriers to curation and making it easier to use automation. I'm also seeing a big push in iGEM, where we've been spreading the word on measurement and engineering methods. And a lot of folks I talk to in government, industry, and around the world all seem to be seeing similar needs and trends, so I hope we're building momentum toward a phase change, and I'm going to see if I can do my part to help.

Check out the full details of our discussion of autonomy in our open access article.