The Lab of the Future will fuse automation, speed and remote work says Thomas Marshall, Senior Staff Research Scientist at Beckman Coulter Life Sciences.
Labs across research areas—life sciences, pharma, and biotech—were already becoming remarkably automated but the pandemic has spurred a new kind of urgency in automating both individual instruments and entire workflows. The need to bring drugs and vaccines to market faster played into the acceleration, but so did social pressures, like the need to work from home and carry out research remotely. Even the public has developed a greater appreciation for automation, if in the form of the need for quick lab results: after all, receiving a negative PCR test result overnight via text or email can dictate whether one can socialise and travel.
We’re certainly not out of the woods yet, as delta and other variants will continue to put pressure on Covid-19 testing and on balancing this lab demand along with existing medical and research needs. Additionally, working remotely during the pandemic has become a way of life that’s unlikely to fully reverse over time. Along with the other major benefits of automation—significantly higher throughput, speed, and accuracy—this all points to leaning into automated solutions as we look to the future.
My own research has ranged from the regulation of gene expression in prostate cancer to gene editing using CRISPR technologies. I now develop novel technologies, particularly reagents, for a range of lab instruments from flow cytometers to centrifuges to liquid handlers. I spend my days thinking about what is lacking in workflows, and what new reagents can help solve the problem. For instance, when Covid-19 testing began to affect the supply chain of proteinase K, used in the extraction step in PCR, I developed a novel strategy to keep extraction intact without proteinase K.
But my career also spanned the shift from manual to automated instruments in various areas, and I well remember when the cost-benefit ratio didn’t always favour automation—now it does. For drug discovery, the benefits were clear earlier on, given the strong need to reduce human input (and errors) and increase the testing rate for new compounds. But the benefits of automation continue to expand across research areas, several of which are discussed below.
Faster, smaller and more accurate
One might envision a medical lab of the future as follows: a sample is sent to a lab, a person drops it into a barcoded system, the machine creates a report showing whether a pathogen or a genetic mutation is present or not, and, if necessary, a recommendation for a specialist is made. Some companies are working on exactly this goal and getting very close to making it reality.
But as mentioned, drug companies also continue to automate more and more of their workflows; the question of whether to invest in more automation is generally whether it gets a drug or a vaccine to market faster. The answer is unequivocally ‘yes’. Automation comes down to the ability to produce many small variations, test them, and then pick the best one. We saw this over the past couple of years, where vaccine makers produced loads of prospects and tested their efficacy in early models. For this kind of work, increasing throughput through increased automation only leads to better end product in a shorter timeframe.
Synthetic biology is another area which is benefitting greatly from automation, including the testing of intermediates or end products of the process. You need to track a multitude of experiments, and with as little human involvement as possible—eight samples can be hard for a human to keep tabs on but tracking 1,536 samples is impossible. The goal is to make a machine that can do the work unassisted, and we’re getting closer to this, even in highly complex synthetic biology workflows.
Liquid handlers are prime examples of the highest levels of automation. Not only do they move liquids quickly and with extraordinary precision, but some also offer remarkable programmability that is easily understandable to the user. There’s little point to having an automated workhorse if it’s hard to interact with. Unfortunately, user interface is sometimes overlooked in instrument design—the question designers should be asking is, how easy is it to set up the system and to reprogram it when research needs change? The ability to set up bespoke workflows is critical, but so is the ability to actually use the instrument; luckily, some of the most highly automated liquid handlers now offer interfaces that are extremely user-friendly.
Acoustic liquid handling is another advanced strategy, particularly for moving liquids in complex patterns. Using soundwaves, these handlers can move a specific amount of liquid on the nanoliter scale from any source well to any destination plate, and so are useful in drug development and synthetic biology alike, among other applications. While the ability to miniaturise workflows can save costs, another beauty of acoustic liquid handling is that it can save time and produces extraordinary reproducibility.
For any highly automated instrument, all the relevant elements work together: If you increase automation and boost sensitivity, you can use less sample and reagent, create less hazardous waste, do the work faster, and arrive at more accurate results. Some aspects do of course become less feasible at smaller volumes—for instance magnetic mixing can work in 384, but not in 1,536, wells—so you have to devise new ways. In my field, we’re constantly trying to increase throughput without increasing footprint. Fortunately, most problems are solvable, and come down to innovating new technology or repurposing what you already have in creative ways.
Shifting roles, creating jobs
While people have worried for decades about robots replacing people, there’s little evidence that this will happen, life sciences industry included. Automation in the lab is clearly shifting people’s time to better endeavours and saving researchers great amounts of ‘grunt work’. The difference between how time is spent by students or researchers 20 years ago vs. now is huge. Depending on the area of research, automation can save days, months, or years of work.
We can move away from people watching machines or acting like a robot to feed devices with sophisticated instruments that notify the user remotely if something is even slightly off. The idea, of course, is not to reduce jobs as we reduce human intervention, it’s about how to repurpose human work into other areas: application development, workflow development, and other areas where there will always be a requirement for human expertise.
In my line of work, the needs of the drug company can often impact what we create—and the need we hear over and over again is more automation, higher throughput, and no touchpoints. As automation becomes more and more advanced, researchers who use automation will continue to gain more time for planning, ideation and development, which will only benefit those who invest in the automation. From what we’ve seen so far, the industry is hiring staff, not laying them off. Automation is about improving processes so that we can do more with the time and skill we have at our disposal.
Running experiments from home
As we continue to move through the acute phase of the pandemic, companies inside and outside of the life sciences have changed their work-from-home policies, for safety reasons and because employees may prefer it. In many areas, including life sciences, organisations are actively keeping people away from work to reduce the spread of Covid-19, particularly as more contagious variants arise. My own philosophy for lab members is to come in to do necessary work and do paperwork at home—there’s no point in hanging around a lab, if it raises one’s risk even minimally.
Luckily, automation makes such a practice more and more feasible. There are now instruments where you literally set up the deck and walk away—the machine even verifies the correctness of the setup using optical technology. Then you can begin and monitor the entire experiment remotely, then come back into the lab the next day if needed. Again, this frees up time for other endeavours, like mapping out experiments down the road.
The reality is that we may never fully get out of Covid-19; even after the surges end, it will likely be endemic and pose other challenges. It’s changed our culture in ways we may not ever fully shake. So if you don’t have to have groups of people in one place, you shouldn’t. Shifting to automation means that one person can be running an experiment remotely and another writing the next phase of the work. This new way of life and work will eliminate human error, improve efficiency and reproducibility, and benefit the safety and health of employees.
Looking forward
These are some of the important considerations we need to be thinking about, as we shift to a more automated world. Most companies are already figuring out how to move into the future if they haven’t yet fully embraced it. And again, we’ll go in this automated direction regardless of Covid—we were moving there already, with more genetic testing and more synthetic biology, areas that have and will benefit from automation the most. Drug developers have long benefited from high throughput, automation, and speed, and will continue to do so. Luckily, we have great minds working on new methods, listening to the needs of researchers, and collaborating on solutions that will pave the way for even more exciting advances in the coming years.
About the author
Tom Marshall is a Senior Staff Research Scientist at Beckman Coulter Life Sciences. He has a PhD in Cell and Developmental Biology from the University of North Carolina. During his tenure at Beckman Coulter Life Sciences, he has managed the Genomics Development team and is currently working on research projects for new product development.
