An argument for increased automation in the life sciences industry
03/01/2004
The myriad of highly detailed and repetitive process steps should be enough to justify pursuing an automated cleanroom solution. Now, how do you do it?
By Jeffrey Walter
The life sciences industry, which was once primarily focused on research and development, is going through a transformation. Institutions, methods and technologies are forcing companies to incorporate higher throughput and more efficient production methods.
These trends have necessitated a movement away from manually operated processes and toward reliable, automated platforms and integrated systems.
As the needs of the life sciences industry continue to change and adapt to new technology and processes, lessons learned from industries with similar needs must be applied. There's little doubt, for example, that the pharmaceutical/biotechnology markets will be able to make great strides based on the electronics industry's knowledge of cleanroom and environmental controls.
Similarities are strong, including the need for a clean and controlled production environment, and a trend toward miniaturization and production. By using a systematic approach for implementing automated solutions, reliable and efficient automated systems can be integrated to replace many manually-operated cleanroom processes in the life sciences industry.
Differences and commonalities
The environmental requirements of life sciences applications are somewhat different from those of the electronics industry. The goal in life science is to reduce the variability of environmental conditions and to provide the most optimal conditions for a particular biological process while maintaining an aseptic environment. The electronic industry's goal, on the other hand, is a significantly lower particle concentration.
For life sciences, concern for mechanical interference of biologically inert particles is less problematic than the contaminant's "living" characteristics. The primary driving force behind maintaining a contamination-free environment for the biological process is to prohibit introduction of bioburden, or a separate, self-replicating contaminant into the environment. This type of contamination may have no effect, may change the course of the process, or may cause the process to fail altogether.
Process enclosures in life science facilities incorporate common cleanroom features and methods, such as HEPA filtration, electrically-enhanced filtration (EEF) treatment, the segregation of critical and controlled areas, pressure differentials, and temperature and humidity control.
Additional measures of control more specific to biological processes may include incorporating an inert gas atmosphere, sterilization by irradiation, ultraviolet lighting, chemical sterilization washes and vaporous hydrogen peroxide (VHP) fuming. While an inert atmosphere may facilitate a process, these other methods are designed to destroy any living contamination within the system and are harsh on the materials and process equipment.
An effective combination of cleanroom technologies and process automation removes human operators from this environment altogether while further limiting the risk of accidental release of potentially life-threatening agents commonly transmitted via an aerosol or a lyophilized particle route.
Process and product liability as well as adherence to stringent regulations are further reasons to maintain control using environmental isolation and automated processes. Clinical diagnostic laboratories are especially concerned with guaranteeing the integrity of their processes used to diagnose patient specimens.
The results of an error in process or mistake in data handling could prove disastrous for the company, if not life-threatening for the patient. Every effort to design an efficient process, remove as many process variables as possible, and maintain stringent records of data and process changes greatly reduce the risk of generating an error that may lead to a process failure.
Once an effective level of process control has been established, the flexibility is in place to quickly respond to changes in regulations and maintain a foothold on the leading edge of market conditions.
A well implemented, flexible, automated platform and controlled environment not only offers exceptional control of the existing process and data, but affords the end user the ability to accommodate changes that allow process improvements, new technologies, limited run-time processes, changes in consumables or perishables—or, more importantly, rapid response to changes in regulations, codes and quality requirements.
Further commonality between the electronics and life sciences industries can be found in the level to which automation has penetrated each one. Methods and products from both industries that are sensitive to fluctuations inherent in a manual or touch labor-based process require process automation.
The electronics industry has been more successful in incorporating automation on a wide scale. Several factors, however, have delayed widespread automation in the life sciences industry, including its research-driven nature and culture, methods of funding, rapid change in technologies, extensive and outdated regulations, and bad experiences with automation.
The driving force behind the production side of the life sciences industry is the discoveries that result from the large amount of lower volume research and development work. These discoveries are often developed using manual labor to perform processes that require an extreme amount of repetition.
Many modular, off-the-shelf (OTS) automated platforms can perform much of this repetitive work. Tasks such as liquid aspiration and dispense, and moving specimens or carriers from one automated platform to another, alleviate much of the extremely repetitive manual labor.
One drawback to automated systems is that they are often based on current diagnostic or process technology. Although many of the process steps performed by these machines are similar, advances and improvements in methodologies, chemistries and consumables often force many of these OTS platforms into early obsolescence.
It's a significant cost for laboratories to continually update their automation hardware to keep their methodology current. This rapid advancement or change in methodology makes it difficult to introduce into a research laboratory a large-scale automated platform that's based on a fixed method.
Inching toward production-type systems
Nevertheless, individual life science companies, their technology and methods—in fact, the entire industry—will be evolving into one that isn't focused as heavily on research and development but one that's relying more on production-type systems.
But this trend has not gone without growing pains. As these organizations grow and realize an increased need for more reliable, higher throughput automation platforms, the decision-makers responsible for accommodating these changes are feeling ill-equipped.
For all their successes in the biology and chemistry aspects of their businesses, most scientists, entrepreneurs or life sciences facility managers suffer from a lack of awareness and experience of systematic approaches to automation that have been evolving for more than a century.
This lack of understanding has caused the life sciences industry to learn some hard lessons in automation—lessons other industries have already learned. The failures have caused observers to have second thoughts about implementing large-scale automation projects.
Expensive lessons, such as failed attempts at total automation of laboratories from front to back, to the contracting of integrators with no life sciences process knowledge, could have been avoided had existing automation philosophy and methodologies been adhered to.
Another change in the life sciences industry that parallels its reorientation towards production is the availability and method of funding. Many of the established companies implementing large scale automation platforms began with limited operating capital.
In fact, commercialization of a product or technology may not have been the primary motive for research. Such organizations are often funded with small grants, institutional funding, and even personal financing. As a result, large amounts of money were not available for small companies to invest into reliable automated platforms. Only recently has the opportunity for commercialization, along with large amounts of venture capital, offered smaller companies the ability to purchase the necessary automated system early in development.
An additional obstacle to implementing production-geared automated platforms is that the pharmaceutical manufacturing segment of the life sciences industry is heavily burdened with antiquated governmental regulations and codes.
Although under review by the FDA, these regulations force a prohibitive cost on the industry to validate or update new quality-control systems. So, to eliminate the cost of validating a new system geared toward manufacturing, the industry has chosen to use the same methods in manufacturing as it had in research and development.
These inefficient production methods have, in part, resulted in a failure rate for medicines manufactured under the current methodology four orders of magnitude greater than the discard rate for semiconductors manufactured by the electronics industry. The cost of these poor manufacturing practices, to both the company and the consumer, seems obvious.
But with the shift from research to production, modernizing of regulations, and an expensive learning curve under their collective belt, the life sciences industry—with proper funding sources in place—is positioned to enjoy the reliability, throughput, flexibility, expandability and return on investment that reliable automated platforms offer.
A need for automation
The myriad of highly detailed and repetitive process steps by themselves would be enough to justify pursuing an automated solution. Strict quality standards, environmental control requirements, environmental and personal safety concerns, federal codes and regulations, as well as operating considerations similar to the electronics industry clarify the need for the controlled implementation of automated solutions.
Using advanced automation products offers the life sciences industry cost-effective capabilities to improve process repeatability, consistency and quality. Automated platforms offer an increase in quality by guaranteeing process uniformity over the incapability of human labor to perform repeated detailed tasks.
In accordance with federal codes and regulations, production and performance criteria can be monitored and adjusted as necessary. An automated process provides improved system diagnostics through real-time equipment feedback. Automation lets a facility increase quality through more thorough control, and offers record keeping that is easily modified to adjust for new requirements, or easily manipulated to generate information.
Many unsafe manual-operated functions can be automated. Extremely repetitive tasks prone to error and injury can be reduced. Exposure to dangerous pathogens and chemicals are eliminated. Moving these operations into a controlled environment with automated systems reduces the risk of injury, spill or accidental exposure.
Most of the benefits of an automated process will be realized by an increase in productivity. An effectively automated process will allow an increase in throughput with minimum modification to the remainder of the process. The process will benefit from an increase in the uptime of the processing equipment; additionally, a reduction in the need for operators will allow reallocation of personnel to a more value-added position.
This is especially true in the life sciences industry where the labor pool is generally well educated, translating to a more expensive labor market. Supervision and efforts associated with employee administrative cost are, in turn, reduced.
Currently, the majority of OTS automation platforms offered to the life sciences industry are lacking in reliability, throughput, functionality and serviceability. Increased reliability, however, is achieved by using automation equipment designed specifically for use in a production setting. Using best-in-class equipment supplied by large manufacturers with national distribution and support services is an excellent method to help guarantee reliability.
Although automation offers obvious advantages to any segment of the life sciences industry, there are still perceived disadvantages. For example, more often than not, current OTS automation platforms common to the life sciences industry tasked with replacing a manual operator still require the attendance of an operator to monitor the system. Also, issues of versatility and flexibility in handling the various consumables or perishables, or performing unusual tasks common to life sciences processes, give the impression that a single automation platform matched with a unique need doesn't exist.
The predominant solution in the industry is to purchase and attempt to integrate many separate pieces of automation equipment to perform the necessary process.
But this solution is neither cost-effective nor reliable.
Implementation
With a need established and a firm understanding of the benefits that automation provides, a plan needs to be put in place that will identify areas in your life sciences facility that would benefit from automation.
A good first step is to define system expectations, such as how to meet specific process requirements, throughput needs and expected improvements from automation.
To achieve optimal benefits, the remainder of facility operations should be evaluated to see how those procedures can take advantage of the automated systems.
You'll need to analyze the operating procedures and peripheral processes to automation to ensure the overall success of automating any process. Significantly increasing throughput in one area of your facility or process may cause automation problems elsewhere, or place a sudden strain on inventory control used to support the newly automated process.
Generate a functional description with sufficient information to document and communicate process needs and performance requirements. Circulate it throughout the organization to generate discussion and firmly establish system requirements.
Seek comments from end users; input from those performing the current process is useful as well. The description should detail general requirements, process steps, safety considerations, data and communication requirements. After review and approval of the functional description, generate a detailed system specification, which should consist of a complete design of the automation system.
As the functional description describes the system's external behavior, the design details the internal function and description of each subsystem. Your detailed design should provide hardware and software requirements, architecture, power requirements, throughput requirements, error recovery needs and detailed system performance expectations.
The value of an integrator
Some life sciences facilities may not maintain employees with the skill sets required to develop a systems specification. If so, a systems integrator who's knowledgeable in the process of the life sciences industr, can assist in developing the detailed specification. An integrator can add value to the project, since considerable engineering resources and integration experience can be applied during project design.
With a reasonable specification in hand, familiarize yourself with standard equipment or companies that are able to design and integrate turnkey solutions.
Careful and equitable consideration should be given to the choice of an automation partner. Considerations should include process knowledge specific to your industry, style of automation, engineering resources, location, training capability, commitment to service and financial situation. The integration partner should apply their knowledge and experience to develop a preliminary systems design for review by the customer.
The customer and integrator will work together to develop a final design review that should detail every aspect of the proposed system, from mechanical drawings and equipment layout to communication protocols.
System construction, control and software development should be transparent to the customer. Integrators should use their life sciences process knowledge, engineering experience, internal quality control and automation tools to create a system that adheres to the process specifications laid out by the customer.
The integrator and customer should work together to generate an Acceptance Test Plan based on the final design documentation. This should consist of a set of testing criteria to verify that the completed system performs to expectations.
Ideally, there should be two acceptance tests. The first should be at the integrator's facility. After the system is constructed, a comprehensive test should be performed to demonstrate the functionality of all hardware and software components. The second acceptance test should take place at the customer's facility after the installation and integration are complete.
Most system components will be covered under a manufacturer's warranty; however, it's imperative that the integrator stands by its work. The warranty should cover software and system controls. Detailed expectations of service and preventive maintenance plans or schedules, as well as operator's manuals and training schedules, should be included in the system price.
JEFFREY WALTER is accounts manager for Biomation, Life Sciences Group of CTA, Inc. (Madison, Alabama). He can be reached at: [email protected]