Justifying laboratory automation
Authored by: Steven D. Hamilton, Hamilton Consulting Group
Data from the 2006 ALA Survey on Industrial Laboratory Automation  indicates, as have past surveys, that improving productivity is the leading reason that organizations implement laboratory automation. There are many definitions and concepts of productivity. Productivity in economics refers to measures of output from production processes, per unit of input. Productivity may also be described as a measure of the technical or engineering efficiency of production. In 1967, Jorgenson and Griliches  defined productivity in a way that may best suit the laboratory environment. According to their formula, changes in input and output have to be measured inclusive of both quantitative and qualitative changes. In an industrial laboratory, quantative changes may be characterized in terms of the cost of headcount, equipment, consumables and reagents. Qualitative changes may include "softer" factors, such as result quality, audit trails, safety, and decision making. This article will describe an approach for evaluating productivity changes that may result from implementing laboratory automation. This approach, originally proposed by Zymark corporation, and later modified, is a simple, go/no-go tool for use prior to investing time in a more detailed justification analysis. A more extensive treatment of the broader economic justification topic can be found in these references.   
Since laboratory automation offers an alternative approach to performing the same procedure manually, the most straight forward quantative way to look at the change automation may impose is via the cost difference of the two approaches. This is a form of Return on Investment (ROI) calculation, which will be familiar to corporate finance staff. For the purposes of this calculation, we will assume that the sample load will remain constant for either approach, as will the cost of consumables and chemicals and reagent preparation time required for the procedure. We will not calculate the ROI for a fixed period of time, as is often done, but the result will be interpretable for various periods of time. ROI calculations are easily manipulated, so care must be taken to input the best possible, impartial data.
Manual staff cost
Calculate the annual staff cost of performing the procedure manually by multiplying the annual staff hours required to perform the procedure by the fully loaded cost of such staff. The fully loaded cost of staff includes the cost of the organizational overhead necessary to support that staff, e.g. facilities, benefits, training, support staff, etc. This figure can usually be obtained from a human resources or finance resource.
Manual process staff cost = annual manual hours * hourly loaded staff cost.
Automated staff cost
Calculate the annual staff cost of performing the automated procedure by multiplying the annual staff hours estimated to be required to perform and maintain the automated procedure by the fully loaded cost of such staff. The fully loaded cost of staff includes the cost of the organizational overhead necessary to support that staff, but may be different from the staff cost figure used above due to differences in the level of staff involved.
Automated process staff cost = (annual estimated automated staff hours * hourly loaded staff cost)
This figure represents the cost of installing the automation, including capital purchase cost, upkeep costs (e.g. service contracts) and costs of any physical plant changes to accomodate the project (i.e. lab modifications). Add to this the staff costs of planning and implementing the project, which should include possible costs to other components of the organization (e.g. database development necessitated by the new system). Again, take care to assign the correct loaded staff cost.
Install cost = Cost of purchase + annual service + physical plant changes + (project planning/implementation hours * hourly loaded staff cost)
A final quantitative figure is calculated by determining the annual staff costs saved by implementing the automated system (manual minus automated) and dividing that figure by the install cost. We refer to this quantitative total as the Economic Justification Index (EJI).
EJI = ((Manual process staff cost) - (Automated process staff cost)) / (Install cost)
An EJI equal to one roughly indicates that staff time savings will pay back the cost of installing the system in one year. An EJI of 0.5 equates to a payback period of 2 years. What an organization does with the staff time savings varies. Some organizations, such as hospital labs, that are strictly regarded as cost centers may choose to reduce staff. Other operations that are more output focused, such as R&D labs, may look at the staff savings as an opportunity to reallocate resources to different tasks or to simply absorb a growing workload.
Although difficult to quantify and therefore not usually included in traditional ROI calculations, qualitative changes can be as, if not more important in evaluating the benefit of an automated system. They can impact the achievement of organizational goals. The qualitative change factors evaluated using this method are:
- Quality: Improved end product quality. The "product" may be data, a purified compound, a cell culture, etc.
- Safety: Isolating people from hazards or isolating the process from hazards or contamination.
- Procedure Enhancement: A resulting end product with attributes that exceed what was produced or possible to produce manually. This could include such examples as: 1) Increased density or resolution of data; 2) Evaluation of more experimental parameters; 3) Conducting a process is is manually impractical, such as creating high-density microarrays.
- Audit trail: A permanent, computer-generated detailed record of process events and results.
- More timely decisions: Improved availability or interpretability of process results leading to quicker decisions.
- Flexibility: The retention of manipulative skills (in the automation) across staff changes and across labs. The ability to rotate through processes/procedures with no manipulative relearning.
The quantitative summary of these qualitative changes we will refer to as the Strategic Justification Index (SJI). Based on the anticipated performance of the proposed automated system and the organizational goals, assign each change factor a numerical value as being:
- Zero: Not important to the organization or applicable to the proposed system.
- One: Beneficial but not compelling to the organization or moderately applicable to the proposed system.
- Two: Very important to the organization and highly applicable to the proposed system.
Divide the sum of the values assigned to the six change factors by a possible high score of 12 to calculate the SJI.
SJI = (sum of six change factor assigned values) / 12
A SJI of 1, for instance, would be an extremely high value, indicating that all six qualitative change factors were highly importatant to the organization and were also highly applicable to the proposed system.
Overall project justification
The resultant EJI and SJI values should be plotted on an x-y graph, with both axes having a minimum of zero and a maximum of one. This assumes that one will not often encounter a EJI>1, but if that occurs, carefully recheck the math, and if it holds up then stop wasting time and go do the project!. In the graph below you can see zones denoting various levels of project justification. As stated at the beginning of the article, this is a simple rule-of-thumb method for quickly evaluating project justification. It is useful as a go/no-go tool before investing considerable time in a more detailed justification tailored to the style and practice of a specific organization. For instance, were one to estimate that a proposed project had an SJI and EJI of 0.2, it is unlikely to be worth investing more time to convince organization management of the merit of the project.
Typically a strong Economic justification is more compelling than a strong Strategic justification, simply because it is more tangible. If the strength of justification is to be built mainly on the Strategic side, it will be necessary to assess in detail the importance of the Strategic Change Factors to the organization and obtain buy-in from key organization decision makers.
Long term considerations
One of the factors that comes into play when estimating costs and benefits of implementing an automation system is the expected lifetime of such a system. Any quantitative payback will accure each year. However, that quantitative payback may erode as equipment ages and requires more maintenance. Qualitative payback may also erode as the equipment becomes less state-of-the-art. Often in dynamic environments, such as R&D labs, it is necessary to spend money updating or modifying automation equipment to keep it useful for changing applications. While it is difficult to accurately predict all these scenarios, they should be given consideration in the planning process. How long do laboratories typically use automation systems before they need significant updates or changes or become outdated? The 2006 ALA Survey asked that question:
|Lifecycle period||% Response|
- ↑ Hamilton, S.D.; 2006 ALA Survey on Industrial Laboratory Automation; Journal of the Association for Laboratory Automation 2007, 12, 239-246
- ↑ Jorgenson, Z. and Griliches D. The Explanation of Productivity Change, The Review of Economic Studies, 1967 34, p249-284
- ↑ Laboratory Robotics Handbook, Zymark Corp., 1988
- ↑ Hamilton, S.D., Kramer, G.W. and Russo, M.F., An Introduction to Laboratory Automation, a short course presented annually at the LabAutomation conference.
- ↑ Principles of Engineering Economic Analysis”, John A. White, Marvin H. Agee, Kenneth E. Case, John Wiley &amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp; Sons
- ↑ Kendrick, J.W. (1984). Improving Company Productivity. The Johns Hopkins University Press
- ↑ Kendrick, J.; Creamer, D. (1965). "Measuring Company Productivity: A handbook with Case Studies" (No. 89). The National Industry Productivity Board
- ↑ Brayton, G.N. (1983 Feb.). "Simplified Method of Measuring Productivity Identifies Opportunities for Increasing It". Industrial Engineering
- ↑ Gurevitch, D.,Economic Justification of Laboratory Automation, JALA 2004;9:33–43.
|Click [+] for other articles on||The Market Place for Lab Automation & Screening||The Market Place|