How automation is fuelling self-driving labs


James Love, Vice President of Automation and Process Optimisation at Novo Nordisk, discusses how automation can help drive efficiencies in labs and how the company has worked alongside Standardization in Lab Automation (SiLA) to improve processes.

As a large pharmaceutical company, responsible for around 60,000 employees worldwide and well-known for supplying almost 50% of the world’s insulin, there are a lot of activities going on in our business. Automation is one such process going on across the value chain. 

In our Automation and Process Optimisation department, we customise robotics, help integrate existing lab equipment to help source the correct solutions , and work out what’s the right technology or approach to solve problems. 

We automate to digitalise, it’s not just about labour saving; it’s about creating really great quality data sets, training people to use equipment such as liquid handlers, and designing new accessories or devices if they don’t currently exist. 

We apply this work to generate data sets for artificial intelligence (AI) and machine learning (ML), seeing then how we can apply it to different experiments going on across the company. 

When we do this, we often look at the different variety of equipment that’s out there. We often use cobots, industrial equipment and a variety of different components and accessories. And in a lot of work, we use the not-for-profit organisation SiLA as our integration tool. We’re a big fan of the open source nature of SiLA, and you can see some of the drivers and servers we have been writing, that we have made open source. We use open sources then because we think this is a good way to get some standardisation across the community and drive things forwards.

An example of how we use this is in process modelling and parameter optimisation. Say you’ve got some input parameters and some output parameters. There’s a black box relating them, you don’t know what the function is, but you think there is a function. So how do you determine it, and optimise it to derive a desired outcome?

We like to use Bayesian optimisation or design of experimentation (DOE). If you can do a lot of experiments, then DOE is a great approach, but if you would like to learn a lot from each experiment and build self-driving labs, then Bayesian is a great choice. We have developed algorithms for process improvement, we have robotic automation to carry out the experiment, and we use maths wrapped inside SiLA to do a lot of the thinking.

Our target is to generate autonomous robots and self-driving robots for different processes, in particular ones working with biologic drugs, which is our focus. 

Examples of this include different liquid handlers and plate readers/imagers; we tie them together with SiLA and then we cycle processes around, do an experiment, learn from it, and decide on the next experiment and just repeat, repeat, repeat. This works away over a weekend or a week, without anybody touching it, helping us learn more. 

Actual successful examples have then the optimisation of different plate-based assays, high-performance liquid chromatography (HPLC) separation methods, chemical and enzymatic reactions and we are working now on stem cell differentiation. These methods are really nice because it means that days, if not weeks, of human time (labour and thinking) are now being carried by robots. SiLA is really the backbone to all this hardware communication and making it all work.

From DDW Volume 25 – Issue 3, Summer 2024 Read the digital issue

About the author

James LoveDr James Love is the Vice President for Automation and Process Optimization in Digital Science Innovation area at Novo Nordisk, A/S, based in Denmark. He is trained in protein x-ray crystallography where he used high throughput techniques to generate proteins and solve protein structures.


Related Articles

Join FREE today and become a member
of Drug Discovery World

Membership includes:

  • Full access to the website including free and gated premium content in news, articles, business, regulatory, cancer research, intelligence and more.
  • Unlimited App access: current and archived digital issues of DDW magazine with search functionality, special in App only content and links to the latest industry news and information.
  • Weekly e-newsletter, a round-up of the most interesting and pertinent industry news and developments.
  • Whitepapers, eBooks and information from trusted third parties.
Join For Free