Search

The Cloud Lab Revolution

Updated: Jul 19

February 25, 2021|Cloud Laboratories, Science and Technology

Will Transform Life and Materials Science R&D Starting Now! 20/15 Visioneers, Thought Leaders in Science and Technology

An Industry Perspective When a set of nascent ideas are brought together in a burst of convergence and integration, delivering new ways to solve widespread problems in mature industries, opportunities arise to challenge the established order. Sometimes this occurs de novo, but more often there are antecedents that point the way. These antecedents offer clues for those who can read them and can help complacent industry players avoid falling into the “follow the laggard” trap that often allows more nimble competitors to steal a march on their business. In the case of the Cloud Lab Revolution, which has the potential to make many legacy biopharma research and development laboratories obsolete, lessons from both the semiconductor industry and the cloud computing business speak volumes.

Several companies have launched commercial cloud lab capabilities, and we will touch on them below. This market will attract multiple players, just as Amazon Web Services (AWS) stimulated Google, Microsoft, IBM, and others to enter the cloud computing business. To best introduce the concept, however, we will focus on the most comprehensive offering we’ve come across to date, sharing insights into where we see this heading.

Before you read further, invest two-and-a-half minutes watching a short video on the home page of the Emerald Cloud Lab (ECL). This will give you a taste of the tsunami about to hit. After reading this article you might want to listen to our in-depth interview of ECL’s founders on Rock Stars of R&D here. Meanwhile, let 20/15 Visioneers (https://20visioneers15.com) describe what lies ahead as we have been preparing for this moment for years.

Emerald Cloud Lab

ECL is the world’s first automated, integrated, comprehensive, remote control, software programmable, pre-clinical biopharma R&D laboratory. It is a robot factory for doing science incorporating just about every cutting-edge idea thought leaders have been kicking around for years to improve the efficiency, flexibility, accessibility, auditability, scalability, reusability, and reproducibility of biopharma research.

ECL-1 is in South San Francisco, constructed in what was once a post office distribution center. Peek inside and it looks a lot like a semiconductor fab from the nineteen eighties where bunny-suit operators moved trays of wafers between rows of CMOS manufacturing machinery. The breakthrough that made costly chip-making facilities accessible outside vertically integrated vendors like Intel and IBM was a company called Taiwan Semiconductor Manufacturing Company (TSMC). TSMC set itself up as a pure-play factory-for-hire running silicon wafers for a diverse set of customers, from startups to industry behemoths. It spawned a parade of new players like Broadcom, Qualcomm, ARM, Xilinx, and Nvidia that transformed the computer and communications industry, leaving the superannuated Bell System in the dust.

Similarly, ECL is a pure-play R&D laboratory-for-hire running experiments for a diverse set of subscribers that range from basic chemistry to cell biology. It is not a Contract Research Organization (CRO) that designs experiments, though labless CROs are expected to hit the market soon using ECL as their back end. ECL’s mission is to accurately and efficiently execute experiments that its customers design. These customers include major pharmaceutical companies, biotechs, university researchers, and fledgling life science startups.

Every ECL customer gets instant access to tens of millions of dollars of best-in-class instrumentation for the price of a subscription, which can run from a few thousand up to nearly a million dollars per month depending on the capacity and throughput required, as well as the materials consumed. The key to capturing the benefits of a cloud lab is the comprehensive nature of the offering. Mixing and matching isolated cloud lab experiments with experiments performed in a legacy manual lab, transferring samples back and forth across what should be a seamlessly integrated experimental protocol, is like trying to take a trip from New York to San Francisco on a partially completed transcontinental railroad interrupted by canal barge, pony express, and wagon train segments. The need to offer comprehensive coverage with no experiment left behind is why it is so complex and time-consuming to stand up a full-service cloud lab.

Every parameter of every experiment in the ECL is precisely specified under the control of the subscribing scientist via commands issued through an intuitive graphical user interface called the Command Center whose output is compiled into a machine-readable Symbolic Lab Language (SLL) software program embedded in an interactive Electronic Laboratory Notebook (ELN). This minimizes the opportunity for human error, inadvertent omissions, or operator bias while supporting the debugging of complex protocols, all before a single experiment is performed. The same approach enables methodical troubleshooting as well as tightly controlled themes-and-variations afterward.

Every piece of data and metadata resulting from every experiment performed across the hundreds of instruments available online is automatically captured by an overlay sensor array that connects every machine, the movement of materials, and every laboratory action to the hive mind of the lab, called ECL Engine, which also does all the master scheduling and resource allocation. This makes experimental data available for real-time use or subsequent analysis utilizing advanced tools built atop Wolfram Language. Every piece of heavily instrumented equipment is calibrated and maintained by a professional staff under rigorous GLP quality control standards. The entire supply chain of experimental materials, including ordering, stocking, tracking, and billing, is executed by an automated procurement and management engine, including the handling of proprietary samples shipped in from clients. This same system can also be used to prepare detailed cost and schedule estimates, supporting the most accurate grant or project proposals.

Everything described above operates 24x7x365 just like Amazon Web Services (AWS) under a pricing regime in which capacity can be scaled up or down to meet budget constraints, short-term, or long-term business needs, with state-of-the-art partitioning, privacy, and security.

Because this comprehensive laboratory operates inside a fully programmable Von Neumann machine, researchers can not only build sophisticated scripts to automatically execute complex suites of multi-stage experiments but can use AI and Machine Learning to design novel exploratory research paradigms in which the input parameters for one suite of experiments are provided by the output data from the prior suite of experiments. This has already opened new avenues for scientific exploration and high-throughput experimentation. Best of all, cloud lab users will accumulate model-quality data (MQD) that will allow them to run virtual assays in-silico, constructing and refining shared or proprietary simulators such that parametric virtual experiments can be explored as a guide to deciding which physical experiments to conduct. It was exactly these types of sophisticated digital tool stacks and the layers of abstraction they embodied that enabled chip designers to keep up with the furious pace of Moore’s Law.

The days of manual experiments requiring “good hands” are over. Say goodbye to “guess and check” science. The laboratory of the future belongs to a new breed of scientist whose success comes from what they do with their heads, not what they do with their hands.

The Numbers Will Make Any Chief Financial Officer Smile

Big Pharma laboratories that used to require hundreds of expensive employees, delivering results in weeks, will be supplanted by public and private cloud labs that take only dozens of employees to operate, delivering results in hours. Top scientists, unburdened by mundane laboratory management tasks, will become super-scientists, empowered by the vast scope and exponential productivity gains delivered by cloud labs. Redundant laboratory re-work and validation often required during inter-departmental handoffs will disappear as transfers become as seamless as passing a digital object identifier. Regulatory compliance, monitoring, and auditing will effortlessly surpass the FDA’s most rigorous record-keeping requirements.

If history is a guide, such productivity leaps will present adoption challenges to senescent corporate cultures in which growing headcounts are a metric of executive success and a ticket to advancement. Lean labless biotech startups, not saddled with obsolete organizational structures and change-resistant management, will spring up like fabless semiconductor companies and mobile app providers, spared the capital investment of building their own labs, the operating expense of running them, or the uncertainty and opacity of working with legacy CROs. Yet their own staff of super-scientists will deliver state-of-the-art, push-button reproducible data that can be easily vetted and replicated by prospective pharma partners. Venture capitalists will have a field day, their early dollars going ten times further.

Principal Investigators at any university in the world, even tiny colleges bereft of wet labs, will compete for NIH grants with the best-equipped laboratories at Harvard, Stanford, and MIT without having to build and maintain vast plantations of post-docs and graduate students, confident that their top-quality scientific work can faithfully feed the technology transfer pipeline while manual lab work continues suffering from replication challenges. Granting agencies like the National Institutes of Health and the Wellcome Trust will be able to support promising young scientists with a fraction of the research dollars, scientists that will not have spent their best years climbing academia’s greasy pole because the only thing a budding scientist needs to join this revolution is a laptop, not a ten-million-dollar lab.

And this revolution will be global. Somewhere in the world’s most remote regions will be an undiscovered Ramanujan of chemistry or biology, a genius destined to cure cancer, who might otherwise never get access to a world-class laboratory. Concerns about fabrication, falsification, and sloppy science coming out of developing countries can be vastly alleviated when that work is performed in a totally transparent and auditable cloud lab, where no data is ever discarded and there are no means to enter fabricated data or phony images into the system.

What will scientific papers look like as this revolution gains steam? First, they will not be frozen “papers.” Scientific work will be presented via interactive digital portals. The Methods Section will no longer be ambiguous prose but will be executable software, instantly accessible and re-usable by investigators across the hall or around the world. The Results Section will not be a cherry-picked collection of PDFs but will include exhaustive cloud-resident F.A.I.R.datasets. The Discussion Section will include live analytics allowing readers to visualize and explore the data in ways that even the authors may not have considered. Peer reviewers at journals with read-only cloud lab accounts will be able to probe the data and metadata behind every experiment, running their own analyses using the most advanced analytical tools, adding these to truly functional Comments and Corrections sections. As a growing body of clean and accessible scientific data accumulates, the ability to conduct complex meta-studies will soar. The plague of disappearing data will end. What does this mean for the scientific publishing industry, ripe for disruption? We can’t wait to help leading publishers and librarians figure it out.

A New World Awaits

ECL is not alone. Other cloud labs are booting up, offering bits and pieces of the puzzle destined for growth and integration. Partnering opportunities abound, and the big players have yet to stick their toes in the water. Strateos, formed in 2019 from the merger of Transcriptic and 3Scan, provides cloud lab engineering capabilities and continues to develop automated systems for chemistry, biology, and tissue analysis. Their pursuit of this vision is driven by two primary products which are Cloud Lab Services and Command and Control Lab Software. Cloud Lab Services is powered by remote on-demand automated cloud labs that are accessible from a web browser that enables scientists around the world to easily access and control new technologies and drug discovery, cell and gene therapy, and synthetic biology workflows. Command and Control Lab Software leverages the power of the Strateos LabOS to enable organizations to control their own scientific instruments and facilities locally and worldwide, schedule workflows, and manage their scientific data. Lilly appears to be their largest of 3 Biopharma client. Arctoris, founded in Oxford in 2016, has developed a proprietary robotic platform focused on drug discovery research. Over the past 5 years, Arctoris has been developing its capabilities to run entire research processes (from target validation to IND-enabling studies) in end-to-end automation. Their Co-Founder and CEO, Martin-Immanuel Bittner MD DPhil commented: "Scientists spent too much time in the lab, and not enough time developing new hypotheses, discussing results with their peers, and really driving projects forward - with Arctoris, we change that, and provide researchers worldwide with highest quality data, so that they can make the right decisions at the right time, and progress their drug discovery programs towards the clinic. This is even more critical in an ecosystem where more and more drug discovery is AI-powered - necessitating access to large amounts of structured, well-annotated data." The role Arctoris plays in this growing sector is also evidenced by the company's recent partnerships with Insilico Medicine and Syntekabio, two leading AI drug discovery companies. 20/15 Visioneers is working on the change, data, and process management side of the equation, with many R&D organizations, and scientific software and hardware clients like High Res Bio, Integrated Lab Solutions, Chemspeed Technologies, Synthace, ACD/Labs, Sapio Sciences, Dotmatics, Perkin Elmer, Benchling, IDBS, Scilligence, and Collaborative Drug Discovery, Scigilian, and Deltasoft in the ELN, LIMS, and process capture environments.

The impact on equipment vendors will be profound. Newly developed instruments from incumbents, inventors, and academia can be quickly brought to market via incorporation into cloud labs, making novel assays, synthesis, and analysis capabilities instantly available across the entire scientific community. Forward-thinking equipment vendors will offer standard interfaces and APIs making it easier to onboard their free-standing instruments, currently a challenging interfacing process. Clusters of related machines that can exchange samples without the assistance of technicians and further advances in robotics will emerge as they did in the semiconductor industry, eventually leading to the reduction in the number of operators required as higher levels of automation became feasible.

Space does not allow for further elaboration, but a moment’s reflection should convince you that this is just the beginning. Want to learn more? Let 20/15 Visioneers be your guide. We offer unique insights and assistance to help companies harness the burgeoning Cloud Lab Revolution. Don’t let the future leave you behind!

77 views0 comments