There are many reasons to introduce automated liquid handling into a laboratory workflow, from increasing productivity or throughput to improving data quality or process security. Unlike most academic or research institutions, drug discovery laboratories need to follow carefully validated protocols, with rigorous quality control measures in place to guarantee the quality and integrity of data. Automation can be a significant boon for these laboratories, not just in terms of achieving the high throughputs generally required for early-phase drug discovery, but also in adhering to SOPs and GLP (and many other three-letter acronyms). Automation also eliminates the interoperator variability associated with manual processes, as well as minimizing the risk of pipetting and logistical errors.
Even a simple automated system can be a significant investment for a laboratory, both in resources and time, and choosing the right supplier for a laboratory’s needs is an essential first step towards incorporating automation into the laboratory routine. A liquid handling system that does not perfectly fit a workflow can create new bottlenecks or limit the scope of research, becoming an obstacle to fully realizing the potential benefits of automation. If a standard liquid handling workstation cannot fulfill requirements, a bespoke automation solution may be necessary to ensure maximum return on investment. However, this creates new regulatory and quality-control challenges, as any modifications to an off-the-shelf liquid-handling platform will likely invalidate the instrument’s type approval. Even for companies with extensive automation experience, overcoming this issue in-house can be a costly and protracted undertaking, requiring personnel and resources to be diverted from other areas to ensure that the laboratory’s output continues to conform to the necessary standards. A more rapid and cost-effective approach may be to work with a laboratory automation provider that is experienced in developing bespoke solutions. If not developed properly, customized liquid handling systems can be a potential minefield for both the manufacturer and the end-user. Any potential supplier should not only have extensive knowledge of the liquid-handling and automation tasks required for a workflow, but also have a good track record for successful project management and a thorough understanding of the stringent regulatory guidelines and applicable national or international directives that need to be in place.
For maximum benefit, a customized automation project should of course be completed in the shortest possible time frame, as well as on budget, but there are important corners that should not be cut to achieve this. The first step in the development of a successful customized liquid handling system is a thorough understanding of its intended use. This makes the initial consultation vital to ensure that the final system specification is able to meet the customer’s needs and expectations. Within the pharmaceutical and drug discovery sector, there is a general trend towards automated solutions that are capable of performing a single application—such as compound screening or cell culture maintenance—as efficiently as possible, rather than the lower throughput and high flexibility usually required in academic research laboratories. This simplifies system development from both instrument design and regulatory approval perspectives, but requires an even greater understanding of the intended application to ensure optimal performance.
Selection and integration
Drug discovery and pharmaceutical development laboratories generally have a well-defined workflow, so the next consideration is which labware types, third-party devices and specialist handling or processing options are required to match the automated solution to existing upstream and downstream processes. These choices can dictate both the day-to-day utility and future scalability of the system, as poor compatibility could result in the need for time-consuming and labor-intensive reformatting or prevent direct comparison of results between methods. From the customer’s perspective, it may also be necessary to integrate existing modules into the automated workflow to avoid the need for full revalidation and recertification of laboratory practices. Successful incorporation of a range of disparate modules into a single, cohesive workstation requires a thorough understanding of both the processing step required and the underlying biology of the application. Identifying the crucial steps of the protocol or assay during the design phase will help to avoid introducing bias as a result of automation, as well as to optimize parallel processing without compromising on data quality.
Once the hardware specification and protocol have been outlined, the next challenge is to integrate the software under a single control program, usually the liquid handling workstation’s control software. At present, there is no software standardization within the laboratory automation sector, making direct interface between different manufacturer’s devices problematic. Even for those instruments with generic drivers available, the rapid pace of development of these systems means these drivers are often unreliable due to compatibility issues or conflicts between different software version or updates. There are moves towards addressing this issue through the Standardization in Laboratory Automation consortium, but currently the most reliable solution for direct control of third-party devices through a single software package is still to develop bespoke software drivers for each individual installation. This requires experienced software engineers with both a broad understanding of the market in order to accelerate development, and specific knowledge of the customer’s applications to ensure straightforward operation.
Issues to consider
From a regulatory perspective, there are a number of areas that must be addressed for any customized system. Firstly, a full assessment of the potential risks of the process—including both hardware and software malfunctions, as well as operator errors—must be made, requiring experienced quality engineers with in-depth knowledge of international laboratory automation legislation and directives. By detailing the differences between the custom installation and a standard, type-approved instrument as part of a failure mode and effects analysis, it is possible to identify potential sources of error, limiting their effects on routine operation of the instrument. This not only aids instrument and process validation once installation is complete, it also helps to guard the laboratory against unforeseen errors that may compromise productivity. Ideally, every bespoke solution should be certified to internationally recognized standards (allowing, for example, CE marking), helping to protect the laboratory from legal issues.
The final piece of the puzzle—once the hardware has been installed, the software drivers have been written and tested, quality control testing has been completed and the individual assay scripts have been developed—is acceptance of the system by the customer. The best way to ensure that the final platform meets expectations is to agree to the functional testing procedure as part of the initial consultation, giving both parties a clear understanding of what the system is expected to achieve. Ideally, this should be a two-step process, with initial testing and approval taking place at the manufacturer’s factory, followed by final acceptance once the workstation has been installed at the customer’s site. This allows any major issues to be addressed before the system leaves the factory, and can save considerable time and additional expense that may have been necessary to correct any issues following installation.
Once an instrument moves into routine operation, customer support is just as important for a bespoke system as a standard installation, if not more so. This makes choosing a manufacturer with a global support network and local presence vital to ensuring smooth running and minimum downtime for either maintenance or repairs. In summary, it is vital to work with a knowledgeable and understanding automation supplier to ensure that bespoke liquid handling solutions perfectly match the individual requirements of each laboratory’s workflow, both now and in the future.
Daniel Moser joined Tecan in 2001 as a project manager for customized solutions, where he held the position of head of project management for the Tecan Integration Group (TIG) from 2005 to 2011 before taking on his current role as marketing manager for TIG. Moser graduated from the University of Zurich in 1996 with a master’s degree in biology, and joined ufamed AG as manufacturing manager, taking responsibility for the adoption of GMP standards in manufacturing processes.