The first step in automating a lab is determining what will ensure present and future process efficiency.
Figure 1. Clinical laboratories are being advised that automation is the
key to survival.
Responding appropriately to economic pressures is key to running a successful laboratory today. Market forces are creating a climate of intense competition that affects IVD laboratories, which are being advised that automation is the key to survival.
Laboratories should be cognizant of the fact that they exist in a business-oriented world and that they need to remain competitive. Challenges to their traditional way of doing things come from a variety of market changes, from cost-cutting tendencies particularly, from the decline of government subsidies, and from a greater awareness of budgetary exigencies on the part of laboratory owners and directors.
When properly implemented, automation systems can reduce overall laboratory expenses, as well as enhance client services and address the concerns of job satisfaction and safety that laboratories face today. The financial savings realized primarily result from reductions in staff size. Increasing productivity with smaller budgets is seen as the way ahead, and automation is central to the strategy for achieving this.
While major IVD companies are offering large-scale integration and total laboratory automation (TLA) for larger laboratories, a trend toward work cell–style automation in medium-sized laboratories, where several medium-sized instruments are linked by a sample-handling system, enables smaller facilities to offer 100% availability of the full suite of tests, albeit at moderate throughput. However, examining and changing processes in order to optimize work flow, rather than merely introducing automation, can be more effective in improving efficiency levels.1
This article explores common misconceptions about automation systems in relation to the size of the companies in which they are installed, and discusses solutions that have been proven to implement a successful automation strategy. It also examines the technical and management considerations required for developing the flexible processes and instruments necessary for a laboratory to automate its systems.
Benefits of Automation
Increasing demands in the area of time to market require that modern IVD laboratories work more rapidly, more accurately, and more efficiently than in the past. Unavoidably, this leads to bottlenecks that threaten the smooth running of the laboratory operation and its efficiency.
Automation has proven to offer many significant production advantages in laboratories, including more-efficient use of personnel, improved system usability (through enhanced operator safety and ergonomics), decreased operating costs, fewer laboratory errors owing to less human intervention, more-rapid processing of samples, faster recording of analytical results, and greater consistency and repeatability exhibited in the acquired information.
The term automation connotes a solution that combines processes into one efficient system that eliminates the risk of errors associated with human performance of the systematized functions. Labor-intensive tasks such as specimen labeling, handling, preparation, and storage that may be executed by numerous staff members, any of whom can introduce mistakes or other problems, can easily be automated.
Labor costs can consume a significant percentage of a laboratory's total budget. By automating crucial yet undemanding tasks, the laboratory can realize considerable cost and labor savings while also seeing large advances in the precision and efficiency of processes.
But before embarking down this obvious path, laboratory companies should ask what level of automation is appropriate for both the current and the future needs of their business? Imaginable future requirements are just as important as current needs, if not more so. Potential future changes in the laboratory operations should be factored into any major new business technology strategy (see Figure 1).
Total Laboratory Automation
Two general types of automation technology now exist, total automation and modular automation. Laboratory automation technology consists of integrated hardware and software able to perform complete specimen processing and analysis. TLA is a large undertaking and suitable for only a small number of the biggest laboratories. As a consequence, modular automation has arisen to provide the benefits of automation to smaller laboratories on a suitable scale while offering scope for expansion and development over a period of time.
In the early days of laboratory automation, TLA was seen as the only solution. Unfortunately, because it was scaled to the largest, highest-volume laboratories, it had limited appeal for smaller facilities owing to its cost. TLA represents a multi-million-dollar investment. It requires a lot of space to implement and so is really available as an option to only the biggest, along with a few midsized, clinical laboratories in the United States.
As its name suggests, TLA automates essentially every process in a laboratory. Its coverage misses only the most esoteric testing. TLA greatly increases capacity. With its implementation, the volume of tests completed rises exponentially, and the cost benefits soar. Expansion potential is assured by the flexibility built into total automated systems. However, although the benefits are obvious and enormous, cost is the overriding factor in any decision regarding whether to introduce automation systems.
Because the benefits of TLA are not disputed, IVD manufacturers have turned their attention to alternative solutions that might be more generally applicable. Modular automation is becoming increasingly popular, as it is a lasting technology that is affordable but delivers the automation advantages of lower costs and increased productivity. Capable of being integrated into existing laboratory production systems, modular systems are compatible also with future operational developments intended to give the laboratory the capacity to provide better-quality services.
Many factors influence a laboratory's decision to introduce automation. Likewise, some challenges to implementation probably will need to be overcome, many of them arising from preexisting inefficiencies. Remodeling of the existing infrastructure, team-building among the personnel, and software interfacing are a few of the issues to be addressed. The enormousness of the task cannot be underestimated. Many laboratories are literally revolutionized by the introduction of automated systems, perhaps sometimes in the bad as well as the good sense of that highly charged term. The changes need to be digested.
Staff reluctance to adopt new technologies—indeed, people's general fear of the new—is a familiar obstacle to the introduction of new systems.
It is recommended that laboratory directors encourage staff participation in the evaluation of existing laboratory processes and try, if possible, to gain staff interest and confidence in the changes that will be introduced. The disruption associated with the introduction of radical changes is a familiar phenomenon. However, it is crucial also to recognize and plan for the scale of the transformation that successful automation implementation can bring to the laboratory.
Few laboratories are suitable candidates for immediate total automation. And many laboratories have considered themselves to be unsuitable for such schemes, their small size being one of the obstacles. Smaller labs do not have the technical resources, finances, physical layout, or big enough market to justify the expense of the transformative procedure. On the other hand, a lack of staff people trained sufficiently to handle challenging tasks may actually motivate laboratories to surmount their other perceived hurdles and employ automation solutions in order to eliminate the capability deficit.
Modular systems have been developed to address these issues, and can accommodate a wide variety of situations. Many small laboratories are being persuaded that flexible yet powerful new automation systems can be valuable to them, seeing them as the way forward in a competitive marketplace. To determine what type of change they need to implement, and how much, laboratories have to evaluate their current systems and analyze the automation options on offer. Flexibility is essential, as future advances must be incorporated and the systems kept current. So is ensuring that the chosen system will be compatible with emerging new technologies.
Meeting the needs of smaller laboratories is the new challenge for TLA suppliers, one characterized by issues of cost, facility size, and staff competency. The large footprint alone of many systems will deter smaller laboratories from selecting them. Smaller, equally efficient machines are available, however. Each of these systems can be customized to suit the target customer's size and needs. The best solution will be apparent once the laboratory has examined its processes to discover which automation approach is most appropriate.
Laboratory managers often neglect to conduct detailed examinations of processes within their facilities. Their responsibility is to run experiments and assays that involve complex and fragile processes that last hours or days. There is usually not time to address the question of laboratory efficiency. Arriving at the recognition that the laboratory needs to change its practices is not such an easy matter, and finding solutions can be even more difficult.
Recently, there has been a move to borrow process models from the engineering and manufacturing industries, a prime example of which is lean manufacturing. At its core, the term lean signifies a fundamental philosophy of eliminating waste in processes of various types. Unfortunately, the concept of waste is often erroneously simplified to mean nothing more than rejected product. “Waste” should instead be taken to encompass all non-value-adding activities, including overproduction, rejects, unnecessary product handling, equipment downtime, and so on.
The idea of lean manufacturing originated in the post–World War II automotive industry in Japan. To survive in very competitive domestic and international markets, Toyota Motor Corp. needed to improve productivity and flexibility significantly without having to invest heavily in automation. The Toyota Production System (TPS) was conceived and progressively refined by the founder of Toyota, Sakichi Toyoda, his son Kiichiro Toyoda, and the company engineer Taiichi Ohno.2 They drew a great deal from the work of statistical process control guru W. Edwards Deming and the writings of Henry Ford. After visiting two vastly different business environments—Ford's manufacturing facility and a U.S. supermarket—for the purposes of study, they created an alternative manufacturing model that had the ultimate goal of eliminating waste, in the sense of wasteful processes. Toyota succeeded in reducing its lead times and costs by using the TPS model, and improved quality at the same time.
Due to its effectiveness, this Japanese lean philosophy has been steadily enhanced, copied, and adapted to other industries. Every company should develop a model that is customized to suit its individual needs. The Danaher group of companies, for example, comprises more than 400 separate businesses, of which the author's company is one. As diverse as these businesses are, at the core of the group's success in highly competitive industries is the Danaher Business System. This has been modeled on the TPS, but is adapted and refined to match the requirements of each Danaher business. Several healthcare companies also have set up internal departments to carry out the specific task of implementing “lean.”
Tools for Evaluating Processes
The essence of the lean concept is not the elimination of processes but the elimination of the waste of any resource—for example, labor, raw material, or equipment. With respect to automation specifically, automated equipment must first and foremost be productive. It is also important to note that the suitability of implementing lean is not scale dependent—it is worthwhile for a company of any size—but the possible solutions appropriate for introduction into the facility will be.
Three powerful tools that enable management to make informed decisions about when and where to invest time and money on a production process are available. They are value stream mapping (VSM), overall equipment efficiency (OEE) calculation, and cost modeling.
Figure 2. (click to enlarge) Value stream mapping is one of the fundamental tools of the TPS. It identifies which parts of the business add value to the end product or service being produced, and which processes add waste (nonvalue) and can therefore
Value Stream Mapping. VSM is one of the fundamental TPS tools.2 It identifies which parts of the business add value to the end product or service being produced, and which processes add waste, or nonvalue, and can therefore be removed. The model maps a production path from door to door, breaking down each task into descriptions of the process or cycle time, downtime, setup time, scrap rate, work in progress, number of employees, and other aspects. It examines each analyzed task for its value, and generates a conception of the future state of the operation that depicts a lean process with waste removed. When applying the VSM tool, companies are often amazed at the amount of waste that is built into an existing process (see Figure 2).
Overall Equipment Efficiency. OEE is a mathematical representation of the efficiency of a particular item of equipment. This calculation tool takes a holistic view of efficiency, comparing actual yield, uptime, and equipment throughput rate with what would be possible under perfect operating conditions. An OEE calculation above 85% is considered best practice in many industries. Given the upfront investment automation requires, and the associated ongoing operating costs, an underperforming item of automated equipment clearly is undesirable. But interestingly, in many laboratories in which uptime is poor owing to issues of sample availability or long batch-to-batch changeover times, OEE figures under 20% are common.
One client of the author's company recently requested an additional piece of automated equipment in order to meet customer demand.
On close analysis, however, it became apparent that, while performing within specification, the already existing equipment had an OEE calculated at less than 30%. The main factors contributing to this poor performance were quickly identified as deficiencies in yield and equipment uptime. The captured data pertaining to each of these factors were used to identify the root causes as variable quality in the incoming raw materials, inadequate operator training, and poor management of production staff. These problems were rectified with minimal investment, with the result that the purchase of an additional machine—and assumption of its operating costs—could be postponed.
Cost Modeling and Scale-Up Planning. Understanding where the costs lie in various processes, and how these costs will be affected by system changes in the future, also is fundamental to making good investment decisions today. Processes that are ideal candidates for automation share certain attributes. They generally are repetitive, high-throughput activities whose demands on employee time generate high labor cost. In healthcare fields, they often are regulated processes that need to satisfy strict quality requirements.
Automating immediately may not be wise, however, or necessary. Much depends on circumstance. Planning for the future with respect to automation involves facing some challenging questions. For instance, while capacities expected in the future might make automation viable, investment in capital equipment may not make business sense in the present. Also, quantifying the cost of achieving or maintaining quality is usually easy in retrospect when data have become available, but doing so in advance of implementing automated systems is often quite difficult.
A company choosing to outsource to a low-cost-labor region must figure out what additional costs entailed in the decision need to be considered, such as tariffs, shipping, and the like. Whether the process in the offshore facility should be manual, semiautomated, or fully automatic is another question requiring attention. Finally, a facility analysis might uncover the potential for legacy equipment or some other process to be utilized instead of automation to fill the need.
Figure 3. By identifying critical processes and selectively automating them, automation technology can be integrated into large and small laboratory systems with ease.
Through the creation of a comprehensive cost model of the process, different scenarios can be explored in detail, and a clear path can be defined on the basis of the level of certainty of assumptions and the company's capacity to invest in the future (see Figure 3).
When making significant investments, particularly in technology, business managers should ground their decisions in an understanding of the fact that the world is changing and uncertain. What might be true today more than likely will not be true tomorrow.
Investing in flexible technology thus can often be an attractive solution. An example is provided by another of the author's company's clients, who developed a high-information secondary screening instrument employing cellular dielectric spectroscopy. The platform is a fully automated system for secondary screening laboratories and target validation. Significantly, the technology was specifically designed to meet both high- and low-volume requirements, so it can be integrated into both large and small laboratory systems with ease. This feature broadened the potential market for the product.
Unfortunately, flexibility often comes with a short-term price premium. Astute system selection and implementation, however, allow this initial cost to be offset by an extension in the useful lifetime of the equipment.
Managing capacity scale-up is always challenging. Often, in the absence of a current revenue stream to justify automation, a solution needs to be found that will effectively satisfy both short- and long-term facility requirements, or at least allow for an easy transition. Identifying those processes that are critical and selectively automating them by means of technology that can be integrated with additional processes in the future is one common strategy for satisfying these conflicting requirements. This strategy is particularly prudent in automating regulated processes, for which changes made in the future to accommodate new requirements can be both difficult and time-consuming.
David James is director of manufacturing
innovation at Invetech (Melbourne, Australia).
He can be reached
In evaluating laboratory processes to see whether work flow optimization requires automation, or to see what sort of automation would be best, it is important to look at the future as well as current needs of the facility, and to develop solutions that meet the laboratory's specific requirements. The suitability of automation is not a function of company size but is very situation specific. Different approaches are appropriate in different circumstances.
Ideally, flexibility should be built into any strategy. This involves automating certain processes now in such a way that additional processes can be integrated with them in the future, as required. An experienced team of process engineers should advise on the most appropriate level of automation for a particular laboratory in order to ensure that processes operate at maximum efficiency, whether automated or not.
The lean manufacturing philosophy, which has universal applicability, is about eliminating waste in all of its forms, including excess labor, low yield, inadequate uptime, and overproduction. Laboratories can take a lesson from the engineering and manufacturing industries, where such models as lean production have been applied effectively to increase efficiency.
1. Robert Speziale, “Is Automation Always the Answer?” in Pharmaceutical Technology Europe [online] March 2007 [cited 16 July 2007]; available from Internet: www.ptemag.com/pharmtecheurope/article/articleDetail.jsp?id=418033.
2. Taiichi Ohno, Toyota Production System: Beyond Large-Scale Production (University Park, IL: Productivity Press, 1988).