Human Interactions in Lab Automation

When you think of lab automation, what comes to mind first? Robots? Rails? Instruments? Data? This is what comes to my mind when I think of automation. Lab Automation has a wide range of topics and applications but also great depths in unexplored opportunities. It can be easy to lose track of why automation is important in a lab.

What would you say if I asked, “What problems should automation solve for a lab?” I talked to numerous experts at Ginkgo, and we all landed on two distinct problem spaces:

  • Humans are spending too much time on tasks that can be automated.
  • Certain tasks are too difficult or impossible for a human to perform.

Sounds simple right? Well, notice how I didn’t talk about walk-away time, reliability, flexibility, or throughput. Let me explain. We often think of walk-away time or throughput as excellent outcomes for our automation tools or ways to measure their success. But the question remains, why? Why is throughput or walk-away time so important to us? For instance, if I were a lab user whose role was to execute experiments in automation but my walk-away time was 100%, what would I be able to achieve with that extra time instead? One can argue reducing the number of users executing tasks is more important than walk-away time. Here is where I think the biggest missed opportunity is. When we think of lab automation, we often think of what a robot is capable of achieving when in reality lab automation should be about humans and how technology can augment human capabilities. It is important to think about automation in terms of what it enables humans to do.

At Ginkgo, we took these problem spaces further. We wanted to come up with essential goals for how we should approach automation, all centered around human interaction. We came up with three.

  • Reduce barriers of entry for humans to leverage automation. This goal is centered around the idea that users of automation shouldn’t be expert programmers or engineers. Tools should be designed for the intended user in the lab and speak its language.
  • Reduce friction for humans to navigate across different levels of automation. This goal is centered around allowing users to easily scale across different levels of automation without having to relearn tools or having a burdensome information transfer. Tools should be integrated to encourage scaling up.
  • Increase the probability of successful automated operations executed by humans and robots. This goal is centered around the quality and reliability of our tools. As we all know, things can go wrong in automation but it is important for users to not only trust robots and instruments but also understand errors that arise without having to decipher error codes.

For full disclosure, these essential goals do present a dichotomy between flexibility and usability. Meaning, we all want the most amount of knobs and features on our tools but also have them be easy to use. What if I told you this was possible with automation? We will leave that for a future post.

Lastly, I encourage everyone working on developing lab automation tools to think of automation as the concept to augment human capabilities. We can do this by listening to our user’s needs in the lab first and allowing technological advancement to come second.

(Feature photo of Bioworks 5 by Kris Cheng)

Posted By