Uncovering the Next Chapter in Scientific Workspaces: The Emergence of Automation, Cloud Labs and Digital Twin Labs

The first frontier: digitally enabled laboratories

This horizon describes the direct transition from manual data transcription and second-person verification to automatic data transcription between equipment and the laboratory information- management system (LIMS). Automating data transcription integrates quality control and data sharing with internal and external suppliers—thereby improving visibility and reducing risk. Integration of quality control and data sharing allows for targeted investments that improve the quality of data. Effectively, this minimizes the need for (predominantly redundant) testing of raw materials and accelerates the release of incoming materials.

Digitally enabled labs

Location of quality-control test execution

• >90% testing in labs
• Some limited testing online

Use of data and advanced technologies

• Automated data transcription between equipment and systems
• Real-time data insights and optimized Schedules afforded by advanced data analytics
• 80% paperless lab

New capabilities

• Data engineers and Data scientists
• Advanced IT systems for complex data capture and analytics

Current availability

• 100% available

Twinning is winning

The core feature of digitally enabled labs is the use of advanced real-time data analytics for ongoing process verification, prevention of deviations and optimization of scheduling and managing capacity. Digital-centric labs use digital tools, including smart glasses that explain the standard operating procedures with step-by-step visual instructions on how to execute any given process.

A digital twin can help predict impact before physical changes are made to a laboratory. Such technologies have been available for a few years now, and the time to impact for each case can be as short as three months.

Drug repositioning, the re-application of approved medicines for new indications, has the potential to significantly cut the time and cost of developing new treatments and can be accelerated using digital twins—in the absence of which, there is a lack of a systematic way to identify initial indications.

The predominant sources of productivity improvements include:

1. Elimination of approaching 80% of manual documentation work and the four-eye principle (two-person review process)
2. Automation and optimization of scheduling and planning to improve the utilization of materials, equipment, and laboratory personnel

According to the World Economic Forum,²⁴ lighthouse factories are “the world’s most advanced factories, which are leading the way in the adoption of Fourth Industrial Revolution technologies”.

Digital twins through the lens of innovation

The digital twin phenomenon is making its mark in the R&D sector, leading to a shift in the way organizations approach decision-making. Digital twins have significant implications for innovation, as they can change the nature of value-added novelty in economic and social spheres. Consequently, digital twins are changing the landscape of innovation.

As an innovation process, digital twins enable collaboration throughout the value chain, facilitating the integration and sharing of data, as well as collaborative product development, manufacturing, operation and maintenance.²⁷ The use of digital twins also empowers users to participate in the innovation process through cloud computing and enables collaboration within the company and with external parties. Further, digital twins also allow for real-time monitoring and direct feedback during the design phase, promoting a more efficient and cooperative approach to innovation.

Automation: accelerating the scientific approach

In research and development, pharmaceutical companies employ robotics, along with collaborative or other advanced automation technologies, to carry out repeatable tasks, from sample delivery and preparation to other lab-specific automation techniques. In many cases, high-volume testing can be performed online instead of in the physical lab space. Automated labs also use predictive-maintenance technology to plan for infrequent tasks, such as large-equipment maintenance, which can be performed by lab analysts with remote expert support.

Automated labs

Location of quality-control test execution

• 60%–80% testing in labs
• 20%–40% testing on the shop floor

Use of data and advanced technologies

• Full automation of testing and non-testing processes

New capabilities

• Lab technicians with knowledge of advanced technology
• Advanced automation and robotics engineers

Current availability

• 70%–80% available

Productivity improvements arise from these factors:

Reproducibility

This is a major concern for the research community, both currently and historically, with associated economic implications and undermining of public trust in science. Increasing the use of automation throughout research laboratories presents a proposition to tackle this persistent issue. Automation can assist in improving reproducibility through a reduction in human-induced variability, increasing the rate of data generation, and decreasing contamination.

Traceability

Completely digitally-controlled experimentation eliminates the need for trivial logging of process conditions throughout any given protocol. Eschewing the need to rely on paper-based records, procedures are logged automatically, with high fidelity, in automated robotic experiments. This process is invaluable for troubleshooting and for experiments that seek to understand the order-of-addition effect.

Throughput

Automation operates in a consistent and highly reproducible manner, with the capacity to run for 24 hours a day, five or six days per week. Overall, this can deliver a marked increase and throughput in terms of the number of experiments per unit of time. The analytics performed on the high-fidelity data sets produced by automation also minimize the number of experiments required to answer any given research question.

Robots rooting for stem cells: a smart investment in research

Automation has been explored to improve the efficiency and scalability of stem cell-derived therapies in the transit toward clinical applications. Several automated systems have been developed, including the StemCellFactory,³⁰ StemCellDiscovery,³¹ and AUTOSTEM, to automate the normally manual stages of stem cell cultivation, including seeding, growth, colony selection, passaging, quality assessment, harvesting, and potential differentiation. These systems use complex control algorithms to improve cell yields and quality and have allowed researchers to generate large quantities of cells for research and testing purposes, speeding up the path to clinical application.

Catch a cloud: remote control experimentation

Across the sector, academics, small start-up firms, and big pharma alike are increasingly turning to cloud laboratories as part of a trend to outsource work.

Researchers are already increasingly renting computing resources over the Internet—from big-name providers ranging from Amazon to the likes of Google and Microsoft—for reasons far beyond the need for emergency backup. The cloud provides researchers who need an extra bolster of computing power to rent the additional capacity as and when needed instead of paying for a more permanent hardware solution.

The virtual frontier: a primer on cloud laboratories

Cloud labs are virtual environments that allow users to access and experiment with various technologies, tools, and services in a cloud-based environment. They are often used for training, testing, and development purposes, and can be accessed over the internet using a web browser. Such cloud laboratories enable researchers to effectively outsource all aspects of bench work and run several complex workflows simultaneously, round the clock. The conversion of an experimental protocol to computer code represents an air-tight means of reducing human error and improving reproducibility.

Users can set up and configure virtual machines and servers, and experiment with a range of experimental configurations and scenarios, without dependency on physical hardware. This represents a cost-effective and convenient way to learn about new technology and develop a skill set without incurring costs and time associated with the setup and maintenance of physical infrastructure.

Photo credit: LabCloud

Researchers can configure their cloud environments to suit their requirements. Although cloud computing cannot handle analysis that requires state-of-the-art supercomputers and rapid crosstalk between machines, it is perfectly sized for projects that are too large to be accommodated by the desktop, but too small to justify the purchase of high-performance supercomputing infrastructure. In the digital space, teams can easily collaborate by sharing visual snapshots of data, computing configuration, and software.

Automation in the life sciences is not a novel concept, particularly in molecular biology, where much of the experimental work requires laborious and repetitive transfer of tiny quantities of reagent from one vessel to another. Alongside this, the outsourcing of time-consuming or difficult elements in the experimental flow is not new. However, emerging laboratories are different—they empower anyone with a laptop and sufficient funds to pay and play with a full repertoire of reagents and the full gamut of instrumentation available.

Emerald Cloud Lab (ECL): empowering remote wet-laboratory experiments

In the same way that users can pay for access to a virtual library of digital content—think of streaming services such as Netflix or Spotify—ECL and other cloud labs can provide access to a vast warehouse of equipment without demanding the need to invest in any capital.³³

Researchers can log onto the ECL dashboard, outline their choice of experimentation; modify the equipment to meet their requirements, and make iterative adjustments along the way; and receive live progress updates on their experiments, with the option of watching the process on video. An AI-based expert poses as a highly skilled technician, enabling users to tweak default values and identify issues that can halt experiments.

Some potential issues looming at the horizon include biosecurity or bioterrorism threats—but an appropriate review of scheduled experiments and sophisticated systems that can flag and reject any instances that appear illegal or dangerous represents a viable preventative measure. In addition, digital footprints are far easier to record and monitor versus traditional lab operations.

The approach emulates a condition in which a laboratory is operating 24/7. Due to their high capacity and continuous flow, cloud laboratories provide tremendous potential for scientists to produce large swathes of data without ever having to enter the laboratory. Another issue is one of intellectual freedom—it's possible that the next generation of scientists, while able to build the infrastructure, will not be able to truly understand how to engage with the software or the hardware that generates their data. This opens the door for big private companies to monopolize that information.

Yet, despite these concerns, the appetite for cloud laboratories is ever-growing. Companies like ECL, alongside another San Francisco Bay area company, Strateos, are working with the U.S. research agency DARPA to understand how facilities can improve the efficiency and reproducibility of experiments, with software licensing to enable other institutions to convert their facilities.

Ultimately, while the world makes room for the robot researcher, traditional research scientists are here to stay—or at least, will not be replaced entirely. Instead, the increasing use of automation and artificial intelligence in research could lead to a shift away from traditional methods. The days of sitting at a bench for long hours in a white coat and gloves, with nothing more than the flame of a Bunsen burner for company, are becoming a distant memory.

²³ McKinsey. Digitization, automation, and online testing: Embracing smart quality control. Available from: https://www.mckinsey.com/industries/life-sciences/our-insights/digitization-automation-and-online-testing-embracing-smart-quality-control. Last accessed: January 2023.
²⁴ World Economic Forum. These 10 new 'Lighthouse' factories show the future of manufacturing is here. Available from: https://www.weforum.org/agenda/2020/09/manufacturing-lighthouse-factories-innovation-4ir/. Last accessed: January 2023.
²⁵ Nature biopharmadealmakers. How digital twins of human cells are accelerating drug discovery. Available from: https://www.nature.com/articles/d43747-022-00108-3. Last accessed: January 2023.
²⁶ Parmar R, et al. Building an Organizational Digital Twin. Business Horizons. 2020;63(6):725–736. doi: 10.1016/j.bushor.2020.08.001.
²⁷ Cheng J, et al. DT-II:Digital Twin Enhanced Industrial Internet Reference Framework Towards Smart Manufacturing. Robotics and Computer-Integrated Manufacturing. 2020;62:101881. doi: 10.1016/j.rcim.2019.101881.
²⁸ Lim KYH, et al. 2020. A State-of-the-art Survey of Digital Twin: Techniques, Engineering Product Lifecycle Management and Business Innovation Perspectives. Journal of Intelligent Manufacturing. 2020;31:1313–1337. doi: 10.1007/s10845-019-01512-w.
²⁹ Innovation 4.0 Playbook: Digitalised Research, Development and Innovation in the Chemical Sciences. Available from: https://www.imperial.ac.uk/media/imperial-college/research-centres-and-groups/digifab/2022.04-KTN_Chemistry-Industrial-Biotech-Report-Innovation-Playbook.pdf. Last accessed: January 2023.
³⁰ Doulgkeroglou MN et al. Automation, monitoring, and standardization of cell product manufacturing. Front. Bioeng Biotechnol. 2020;8:811. doi: 10.3389/fbioe.2020.00811.
³¹ Jung et al. Highly modular and generic control software for adaptive cell processing on automated production platforms. Procedia CIRP 2018;72:1245–1250. doi: 10.1016/j.procir.2018.
³² Ochs et al. Advances in automation for the production of clinical-grade mesenchymal stromal cells: the AUTOSTEM Robotic Platform. Cell Gene Ther Insights 2017;3:739–748. doi: 10.18609/cgti.2017.073.
³³ Nature. Cloud labs: where robots do the research. Available from: https://www.nature.com/articles/d41586-022-01618-x. Last accessed January 2023.