What is the Lab of the Future?
The Lighthouse Vision & Backcasting to Improve the Labs of Today
Labs today consist of a series of disconnected machines that are very rarely integrated in any way. While some areas of the lab are extremely advanced, others lag, leaving scientists to spend valuable time and energy focused on low-value tasks when they should be free to analyze and discover the data they collect. The Lab of the Future (LotF) envisions a new lab experience that enhances human processes and connects equipment for a smarter, more efficient experience.
If you ask 10 people what the LotF means to them, you’ll get 10 different answers. Partially because, by definition, the LotF is coming sometime in the future, but also because many are unsure. Most people begin to define the LotF by setting some guardrails, selecting an arbitrary fixed point in the future or describing some nebulous concept of an “agentic lab” or “lights-out automation.” But to arrive at any useful definition for the LotF, we must recall the fundamental objectives that the LotF is supposed to achieve.
Greater Than the Sum of Its Parts
If you think like a scientist, the goal of any study is to understand healthy human biology. A healthy human (or animal, plant, etc.) is the base case from which diseases and illnesses can be interrogated to identify their biomolecular basis, allowing for the design of a safe and efficacious medicine to restore health. To build this beyond-accurate base case, I need three things:
- I need a dynamic data model that brings together sources of biological knowledge.
- I need a functional layer that allows me to interrogate my current working knowledge of human biology.
- I need to be able to update my model as new information is generated.
When considering the LotF in this context, it becomes clear that “automation” or “GenAI” are only features of the LotF, not technologies that define it entirely. These features are only useful insofar as they enable new, validated information to be generated and added to the reference data model.
It then becomes evident that in the LotF, data must be born Findable, Accessible, Interoperable and Reusable (FAIR) to enable a dynamic data model. FAIR data allows us to preserve the context in which that data was generated — meaning it’s durable and can contribute to future analyses, whether it be the next day or decades from now. The use of GenAI or agentic AI in the lab becomes evident as well. No single human scientist or team of scientists could monitor every piece of data or study every novel finding in the model. But an AI agent that continuously interrogates the model and presents a summary of those inferences to a human is the functional layer every scientist needs.
Applying the Thinking
How does this apply on a practical level to pharma and biopharma (and animal health, agriscience, etc.) research? First, we need to separate the LotF into two dependent, but separate, functions that are ripe for transformation: Operations and Analytics. Operations take over the mundane tasks of running a modern biomolecular lab — inventory management, documentation, training, regulatory compliance, etc. Operations technology also can prevent avoidable errors — scheduling, reagent integrity and so on.
Analytics uses metadata from Operations to qualify experimental data, but also enables better experimental design, interpretation and ultimately prediction of experimental outcomes in silico — which can be empirically confirmed (or not) by running the experiment in the lab. Admittedly, our in silico representation of human biology is at some distant (but not too distant!) point in the future.
Here Today, Future Tomorrow
So, what about right now? Well, defining a lighthouse vision and backcasting to a current state is all part of EPAM’s ethos. This consulting process bodes well for concepts like the LotF that are loosely (and variably) defined. In an analysis of the labs of today, we can define wins for the coming years — stepping stones to our dream state. Using human-centered design and ethnographic principles, we can study scientists in a variety of lab settings to identify moments in their operational and analytic workflows that can be transformed with technologies we know and love today … and we have. But that’s a story for a different day.
EPAM Life Sciences is home to many ex-industry scientists with a deep understanding of lab informatics systems and researcher workflows. Coupled with superior delivery teams and experience with emerging technologies like Intelligent Automation and Artificial Intelligence, we’re prepared to help your organization design and build its ideal Lab of the Future.
