Robotics And AI Will Reshape The Life Sciences Buildings Of The Future
Whether it’s the mRNA vaccine technology that powered the Pfizer and Moderna Covid vaccines, new frontiers of biomanufacturing or new personalized medicine enabled by understanding the human genome, medicine is evolving fast. For those working in, designing and building the labs where these discoveries are made, it’s clear the workspaces are changing as well, challenging life sciences developers to adapt to emerging technologies.
A pair of innovations have been steadily making their presence felt in labs: the use of artificial intelligence to sequence large amounts of data and analyze large sets of experiments — and even experiment via advanced simulations — and the use of robotics to run precision testing, verification procedures and advanced manufacturing. AI and robotics are both becoming more affordable and accessible for biotech companies and startups, and becoming the focus of a new breed of startups focused on lab automation.
For life sciences developers working on projects set to open years in the future, industry experts say it is important to get a handle on these technologies now, as they begin to become more mainstream and robotics become smarter, smaller and more efficient.
“They are going to become more accessible, and a few years from now, will be more broadly adopted,” said Facility Logix founder and President Pat Larrabee, whose consulting firm specializes in life sciences buildings and bioscience clusters.
It can take two to three years to design and develop a facility, Larrabee said, and by that time, lab automation is likely to become more prevalent. It removes human error, results in more consistent quality data with everything a company does, and it analyzes different options faster, she said. Labs often aim for Clinical Laboratory Improvement Amendments, or CLIA, certification around consistent data; Theranos, the failed blood analysis startup, had its CLIA certification withdrawn.
“It enables you to take a thousand potential solutions and focus on 100 potential solutions,” Perkins & Will's science and technology principal, Matt Malone, said. “What we’re talking about is speed to market.”
The tech isn’t a common part of lab setups today, Larrabee said, but there is expected to be swift adoption as many newer life sciences technologies mature.
For instance, in the gene therapy field — where personalized medicine means a different product for every patient — labs creating therapeutics may have machines from five or six different vendors on the process line. AI and robotics functions as an integrator of sorts; Larrabee compared it to a master control in the Internet of Things. It’s also expensive, likely for use only for custom-built labs and not in incubator spaces.
Eventually, Larrabee said, this technology will mean less lab space for individual companies. Increasingly, some clinical development R&D could be outsourced.
“In theory, you can run systems overnight, put labs in basement space and make the building run as efficiently as possible,” Malone said. “You can leave the lab and turn the light off.”
Benchling, a company that built a cloud platform specifically designed to accelerate life sciences research and development, counts 300,000 scientists and 1,000 R&D organizations globally as customers, including many focused on new technologies such as CRISPR, CAR-T immunotherapy, RNA therapeutics and gene editing.
The companies see automation, as well as externalization, becoming larger parts of the laboratory ecosystem. Automated labs can work 24/7, and contract research organizations can be used to outsource part of the research process, especially to less-expensive markets, and collaborate more easily across time zones.
“With the rise of new, complex drug modalities, as well as increased competition within the life sciences and biotech space, we’re seeing companies increasing the throughput of their research rapidly,” Benchling Product Manager Tara Lee said. “They are increasing investments into instrumentation and robotics to increase efficiency, and with that, investing in dedicated laboratory areas to house their automated systems, such as automated freezers, automated incubators and liquid handlers.”
Developers would be smart to think of laying out and designing spaces that can be subdivided in the future, Larrabee said, as large labs that today accommodate one client may eventually fit two as space requirements shrink. Guaranteeing that utilities, loading docks and other amenities can be used by two clients if a space was split is a way to hedge against technological disruption.
E4H Environments for Health Architecture Director of Health Sciences and Technology Jeffrey Schantz advises developers to think of their regional ecosystem when considering a spec development utilizing these technologies. How will it fulfill the needs of firms within the region, especially the kinds of biomanufacturing or research likely to spin out of local research centers or graduate from incubation spaces? For those looking to, say, tempt Boston-area startups to move, it’s key to overdeliver amenities.
“Past pilot stage, there’s no reason to be in Kendall Square,” he said, referring to the Cambridge, Massachusetts, neighborhood that is the tightest and most expensive lab market in the country. "You shouldn’t pay that much for rent when you’re manufacturing."
Clinical testing is at the forefront of developing robotics technology in labs, Malone said. Quest Diagnostics, for instance, just built a new testing facility in New Jersey featuring new automation technology.
When designing, Schantz said he will often set up what he calls a "sandbox." It’s not always clear what types of machines will be used in such a setting, but designing a space with sufficient process piping that is protected from electromagnetic field and vibration, can satisfy the requirements of most biomanufacturing spaces.
Developers can plan ahead for the increasing use of robotics and AI in laboratories, Schantz said. It’s crucial to consider utilities and make sure there is enough power capacity for larger, more complex machines.