From wet to dry: How AI is shaking up laboratory design

Laboratories are undergoing some big changes thanks to a common disruptor: artificial intelligence.

Advanced mathematical modelling, computational data analysis and generative design are boosting demand for dry labs, which, as the name implies, differ from so-called wet laboratories, with their use of liquids, chemicals and biological samples.

Wet labs and hands-on science remain core to research, but one global consulting firm estimates that generative AI could produce up to $28 billion of annual value in drug discovery alone.

“We’re already seeing demand for dry labs rise as big pharma organizations look to upgrade infrastructure,” says Richard Cairnes, JLL’s PDS UK and EMEA Head of Life Sciences. “It might be creating facilities in new countries, with scientists collaborating together via the cloud, or simply adapting existing labs to future proof and complement current research resources.”

Take the UK’s Wellcome Genome Campus, where ongoing development includes large open-plan areas for dry lab work and data analysis alongside dedicated spaces for high-performance computing and AI research. Or Germany’s Max Delbrück Center for Molecular Medicine (MDC) in Berlin, which has added substantial dry lab space for bioinformatics and computational biology.

While key breakthroughs will still come from human scientists, Gul Dusi, JLL’s Managing Director for PDS Life Science Projects in the U.S., believes that more extensive use of modelling and AI will fundamentally alter laboratory design.

“It affects the overall layout, altering the number of benches, power, server and data connections required, as well as how people move and interact in the lab space,” she says.

Dry doesn’t mean simple

While dry labs don’t require the same design or infrastructure as wet labs, there are other considerations, such as the need for robust power and HVAC systems to support a higher density of hi-tech equipment.

It means that while repurposing stranded assets to dry laboratories is one possibility, not all buildings are suitable for adaptive reuse.

Dusi highlights quantum computing labs as one example.

“It's one of the most complicated buildings to construct because it demands an almost astronautical like environment with no atmospheric pressure, created by tanks of nitrogen and argon gas,” she explains.

Aside from power requirements, Dusi adds that dry labs may still need substantial load bearing for large or heavy equipment, have deck to ceiling height requirements or vibration considerations.

Cairnes agrees and says that for developers and landlords looking for ROI, the capital outlay for contrasting technical elements will make it harder to fit out dry labs on spec, as tenants will have very specific requirements.

“While the physical building costs may not differ hugely to traditional labs, it’s the more complex AI, automation and robotics equipment needed that will push up the bill,” he says. “It’s likely that the provision of flexible lab space that meets the needs of end user scientists and their specific science plans will remain and be key going forward with the lab of the future.”

Digitization supports faster innovation

Project management professionals are now using digital tools and AI to create time and quality efficiencies for more strategic and cost-effective construction of life sciences projects.

AI’s ability to collect, organize and interpret large volumes of information to extract useful insights can help with everything from procurement planning and program scheduling, to monitoring site safety, or improving sustainability.

Cairnes explains how building information modelling (BIM) helps create digital twins for visualization and better planning. “For example, it can detect potential clashes between pipes, ductwork or electrics and structural elements such as beams, which could cause expensive problems further down the line,” he says.

For Dusi, AI’s potential to enhance the overall experience and wellbeing of people working in life sciences laboratories is what’s most exciting. She sees huge potential for AI to simulate various scenarios and create evidence-based design for greater productivity and efficiency.

“By looking at the path of access for the scientists, how many steps it takes between various bits of equipment, how they interact with their colleagues in both wet and dry labs, as well as things like air quality, daylight, we can design and build labs that help researchers achieve key breakthroughs faster,” she says.