Automated cell culture systems: How automation prevents $100,000 lethal errors

Automated cell culture is viewed as a milestone for scalable, robust, and cost-effective biologics and cell therapy production. As personalized therapies advance, quantitative evidence continues to demonstrate the advantages of semi- and full automation across process robustness, quality control, and replication — with consequences that extend well beyond the lab bench.

Bioprocessing and its automation draw roots far beyond this century, as mass production of living organisms, such as yeast and fungi, experienced unprecedented demand in both research and everyday application. Early bioreactors and their accompanying infrastructure, including automated impellers and pumps, were partly driven by the urgent clinical need for penicillin all the way back in the 1940s (1).

With what many experts now call the fourth industrial revolution, a fully automated facility manufacturing cell and gene therapies (CGTs) is predicted to emerge by 2030 (2). The objective behind such facilities is not entirely financial, as reduction of human error and strengthened process robustness represent systemic requirements enforced by major regulatory bodies, including GMP, FDA, and EMA.

Adopting closed processing and automated cell culture systems can prevent scenarios in which up to 40% of stem cell batches intended for therapeutic use are identified as contaminated — representing both significant financial loss for manufacturers and lost therapeutic opportunities for patients (3). These improvements are well-documented, and the market for automation-driven innovation continues to expand.

This entry focuses on the real-world consequences of implementing automated workflows in bioprocessing, and the risks associated with failing to adopt these approaches in time. While automation is most effectively addressed at the facility planning stage, a strong protocol must also remain flexible, accommodating re-validation and continuous improvement as the field evolves.

Not even established CGT manufacturers can entirely prevent the failure rate of <5%

Contrary to common belief, process robustness remains an unsolved challenge even for leading cell manufacturers. Reports indicate that only approximately 90% of Novartis Kymriah CGTs are successfully shipped and distributed, with manufacturing failures accounting for the shortfall. In this industry, a failure rate of even 5% — arising from contamination, operator error, or equipment malfunction — carries unacceptable consequences at commercial scale, both financially and for patient safety (4).

Eliminating heavily manual operations through automation reduces failure-to-follow-procedure events, which remain among the most frequently cited observations during FDA inspections of GMP facilities. Biological variability, however, is considerably harder to control. In particular, distinguishing between insufficient process robustness and raw material variability remains a persistent and technically nuanced challenge (4).

In pursuit of robust processing, automating key process parameters delivers measurable value across both reliability and efficiency. Automated media exchange and cell confluency assessment, for instance, can be executed with greater accuracy, speed, and consistency than manual equivalents. Online monitoring of morphological and metabolic parameters provides early indicators of phenotypic change throughout the cultivation process. In suspension-based cell cultures, where harvesting protocols are more amenable to automation, certain end-to-end, handling-free platforms are already commercially available (5).

As technological maturity in cell and therapy manufacturing continues to advance alongside the field itself, even modest reductions in human error translate to significant savings. A single failed batch is estimated to cost upwards of USD 100,000 in manufacturing expenditure alone. Consequently, the convergence of automated cell culture systems with the broader adoption of advanced therapies in healthcare will likely strengthen both the accessibility and the commercial incentive to implement them (6).

From “test-and-hope” to quality-by-design: safety and potency continue to improve

Process control in therapeutic manufacturing involving living organisms is inherently challenging, and depends heavily on the disciplined monitoring of critical process parameters (CPPs), which include temperature, CO₂ concentration, nutrient availability, and cultivation time. The relationship between CPPs and a product’s critical quality attributes (CQAs), framed within the quality-by-design (QbD) framework, has become central to the digitalization of biomanufacturing — increasingly powered by machine learning (7).

The QbD approach ensures that CQAs are predetermined and aligned with regulatory requirements before manufacturing begins, acting as a prerequisite for patient safety and responsible resource allocation. Maintaining an optimized process has been shown to reduce therapy development timelines by up to 40% and material waste by up to 50%, providing strong incentive for the development of automated algorithms designed to sustain these gains (8).

Furthermore, the accumulation of large-scale biomanufacturing data has opened a credible pathway for machine learning integration, enabling real-time prediction of CPP deviations and their automated correction as a quality control technique. Limited regulatory approval for machine learning applications in this context, however, represents a tangible barrier to near-term adoption (7).

Unlike the consumer goods sector, where a product recall is a manageable corrective action, complex biological therapeutics offer no equivalent safety net once an error is identified at late manufacturing stages. In the case of chimeric antigen receptor (CAR)-T therapies, failures in end-to-end process control can result in life-threatening adverse reactions, as illustrated by the 2026 ESO-T01 trials (9). Where automation offers a viable mechanism to prevent such outcomes, limited prior experience should not necessarily serve as justification for delaying validation of improved quality control approaches.

Replication is where the manufacturers’ financial interests converge with scientific trustworthiness

Whether culturing human mesenchymal stem cells (hMSCs) or human embryonic kidney (HEK) cell lines, shared bioprocessing principles apply across contexts, ultimately culminating in the requirement for replicable outcomes. Over repeated cycles of a workflow, the manufacturer must reliably deliver consistent cell quantity and product efficacy, independent of the intended downstream application. While this may appear conceptually similar to reproducibility, replication specifically refers to the ability to achieve consistent results across previously unseen data and conditions (10).

The introduction of automation into cell counting — a routine but operationally critical step in most culture workflows — illustrates this principle clearly. A comparative study using Chinese hamster ovary (CHO)-K1 cells and two independent replicates demonstrated that result deviation fell from above 9% with manual methodology to below 6% with a semi-automated approach (11). While this may appear to be a modest improvement in isolation, this extent of reductions in variability carry substantial implications for batch consistency and regulatory compliance at commerical stages.

Further evidence comes from a comparative analysis of process capability estimates (i.e., Cp and Cpk values) across manual and prototype automated workflows applied to human osteosarcoma (HOS) cells. While neither method negatively impacted cell viability, manual handling produced poor process capability (Cp < 1.0), attributed primarily to excess variability. In contrast, the automated cell culture workflow achieved process capability improvements of several orders of magnitude following adjustment of process measurements within defined specification limits (12).

These findings reinforce that replication across culture seeding, incubation, and harvesting directly shapes the entire trajectory from laboratory development to industrial-scale manufacturing. The absence of this fundamental capability does not manifest as a minor technical gap, but rather represents a systemic risk that is both recognized and acutely felt across the scientific community.

The case for automated cell culture systems in bioprocessing is no longer a speculative one. From reducing human error and contamination-driven batch failures, as well as their associated six-figure financial consequences, automation in modern biomanufacturing is meant to address the points of greatest vulnerability.

As the complexity of advanced therapies increases and patient safety expectations by the major regulatory agencies intensify, the margin for process-related error is persistently narrowing. For manufacturers navigating the transition from manual to semi- or fully automated operations, the data tell a consistent story: the cost of inaction will exceed the cost of adaptation. The question is no longer whether to automate, but how to do so compliantly and without having to rebuild the workflow entirely.

From evidence to action: Archimedes® One

The core of Archimedes® One — our adherent dynamic bioreactor — is equipped with a compact 10,000 cm² vessel, engineered for reproducible performance in GMP-compliant environments. Designed to support a broad range of cell types, from primary stem cells to established research lines, the technology simplifies adherent cell production while intensifying process control.

Visit our brochures and literature page or join the Early Access Program to discover how Archimedes® One can reduce human error-driven costs in your workflow.

References:

1. The History of Bioprocessing. (2022). Phill Allen, ALLpaQ Packaging Group. Available https://allpaq.com/the-history-of-bioprocessing/ (Accessed 07 April 2026).
2. Robots in Biomanufacturing: A Road Map for Automation of Biopharmaceutical Operations. (2022). David Wolton & Carl-Helmut Coulon, BioProcess Interational. Available https://www.bioprocessintl.com/information-technology/robots-in-biomanufacturing-a-road-map-for-automation-of-biopharmaceutical-operations (Accessed 07 April 2026).
3. Szabłowska-Gadomska, I, Humięcka, M, Brzezicka, J, Chróścicka, A, Płaczkowska, J, Ołdak, T, Lewandowska-Szumiel, M, Microbiological Aspects of Pharmaceutical Manufacturing of Adipose-Derived Stem Cell-Based Medicinal Products (2023), Cells 12, pp. 1-16. doi: 10.3390/cells12050680.
4. What failure rate are we willing to accept in cell therapy manufacture? Insights from medical device development. (2020). Dr Dan Strange, TTP. Available https://www.ttp.com/insights/what-failure-rate-are-we-willing-to-accept-in-cell-therapy-manufacture-insights-from-medical-device-development-2 (Accessed 10 April 2026).
5. Moutsatsou, P, Ochs, J, Schmitt, RH, Hewitt, CJ, Hanga, MP, Automation in cell and gene therapy manufacturing: from past to future (2019), Biotechnol Lett 41, pp. 1245-53. doi: 10.1007/s10529-019-02732-z.
6. Automation in Cell Therapy Manufacturing. (2016). Ian R Harris, Francis Meacle, Donald Powers, BioProcess International. Available https://www.bioprocessintl.com/cell-therapies/automation-in-cell-therapy-manufacturing (Accessed 10 April 2026).
7. Walsh, I, Myint, M, Nguyen-Khuong, T, Ho, YS, Ng, SK, Lakshmanan, M, Harnessing the potential of machine learning for advancing “Quality by Design” in biomanufacturing (2022), MAbs 14, pp: 2013593-604. doi: 10.1080/19420862.2021.2013593.
8. Rethinking Pharmaceutical Industry with Quality by Design: Application in Research, Development, Manufacturing, and Quality Assurance. (2025). Melissa Hennig, PharmaExcipients AG. Available https://www.pharmaexcipients.com/news/rethinking-pharmaceutical-industry-qbd/ (Accessed 10 April 2026).
9. In Vivo CAR T Causes Serious Toxicities in All Patients of Early Trial. (2026). Christoph Burgstedt, Inside Precision Medicine. Available http://insideprecisionmedicine.com/topics/oncology/in-vivo-car-t-causes-serious-toxicities-in-all-patients-of-early-trial/ (Accessed 10 April 2026).
10. The National Academies of Sciences, Engineering, and Medicine. (2019). Reproducibility and Replicability in Science. Washington, DC: The National Academies Press. doi: 10.17226/25303.
11. Ramm, S, Odefey, U, Frahm, B, Pein-Hackelbusch, M, Semi-automated vs. manual: Comparative study of cell culture counting methods using validation parameters (2024), bioRxiv, pp: 1-19. doi: 10.1101/2024.05.30.596619.
12. Liu, Y, Hourd, P, Chandra, A, Williams, DJ, Human cell culture process capability: a comparison of manual and automated production, J Tissue Eng Regen Med 4, pp: 45-54. doi: 10.1002/term.217.