Managing Risks Related to Custom Automation Projects

Originally Published MDDI June 2002 AUTOMATION Automation equipment for assembly or packaging can benefit from the use of custom systems if builders and customers develop a strategy to share the risks.

Originally Published MDDI June 2002


Automation equipment for assembly or packaging can benefit from the use of custom systems if builders and customers develop a strategy to share the risks.

Lawrence Allingham and François Paillusseau

Risk is inherent in projects involving custom-engineered automated manufacturing equipment. It increases significantly, however, when the project requires technical solutions whose feasibility is unknown at the outset. Moreover, in a dynamic industry like medical device manufacturing, the persistent market demand for innovative products compels firms to implement new and ever more advanced production methods. Thus, a device manufacturer's willingness to undertake innovative automation projects and its ability to manage the associated risks often spell the difference between seizing a competitive edge and suffering a serious setback.

In an innovative machine project, risk besets both customer and supplier. The customer, or end-user, risks the timely introduction of a new product, the realization of a specific return on investment, or the achievement of production goals and requirements. The machine builder typically has even more at stake, because a large and challenging project gone awry could seriously damage the company's financial performance or even threaten its survival. In many cases, custom equipment is specific to a given product, and any innovations developed for a custom project may not be applicable to other machines. This means that the builder will have no opportunity to recoup current losses in the future.

Given these stakes, how can the interests of both the customer and the supplier be satisfied and an innovative machine for automated production be built on time and on budget?


The first step requires the parties to work together as partners—with clear meaning given to this term. In this context, a partnership exists only when both parties recognize that risk is mutual and must be jointly managed for the protection and benefit of each partner. Thus, each party must resist the natural, self-interested tendency to shift risk to the other.

For the end-user, this means recognizing that the machine builder is breaking new ground and may need to test several solutions before finding one that works. It means accepting uncertainty or variability in the machine concept as long as uncertainty or variability exists in either the product or the production process. It means recognizing that the debugging process could reveal some component issues that were impossible to foresee at the outset (e.g., surface-state variability, burrs, static electricity, etc.), even with careful analysis and prototyping. It means respecting the builder's legitimate financial requirements, not only in terms of profit, but also in terms of cash flow.

For the machine builder, this means investing time and effort in the early phases of the project when part designs are fluid and samples tough to come by, and when there is no guarantee that a machine will ever come to fruition. The builder must also avoid the twin temptations of lowballing a price estimate in order to establish a competitive posture and padding the price and lead time as a substitute for proper risk assessment.


It should be emphasized that sharing risk fairly does not necessarily imply sharing risk equally. The flip side of risk is reward, and risk must be shared in proportion to reward. Both parties must understand their respective risk/reward profiles so that they can come to agreement on how risk is to be shared.

For example, if the innovations developed for the machine in question cannot (because of either market or contractual reasons) be applied to other machines for other customers, then the risk/reward ratio for the builder is relatively high. On the other hand, if the innovations could be readily and broadly marketable, then this ratio drops considerably. The flow charts in Figures 1 and 2 illustrate the project phases in high- and low-risk automation development projects.

Figure 1. Traditional machine construction carries a igh risk level for both supplier and customer when the machine is highly innovative.
(click to enlarge)

Likewise for the customer, the consequences of an underperforming system must be identified. Would it result in an irrevocable loss of market share, or would it result in costs that could be recouped elsewhere? The answers to such questions are critical to determining a fair risk-sharing agreement.

The second step is to follow an iterative process of technical validation and price definition. In the early phase of an innovative machine project, numerous unknowns can hinder the development of a particular technical solution. During this phase, the machine builder should not attempt to provide a firm and binding price. A budgetary estimate should be offered instead, based on a preliminary machine concept and expressed within a certain price range, or "envelope."

Figure 2. By conducting one or more phases of preliminary engineering and prototyping, risk can be managed and greatly reduced.
(click to enlarge)

At the same time, however, a firm price should be quoted for a separate, limited phase of research and development designed to study, prototype, and evaluate one or more concepts. The price of this phase of R&D should be a small percentage of the estimated machine price (typically 2–7%, depending on the level of prototyping required). Yet the amount should be sufficient to permit validation of the concept's technical feasibility and narrowing of the machine's price envelope.

For example, a medical device customer was interested in using continuous-motion machine technology to assemble a high-volume, long-running product at rates dramatically higher than those of existing machines. Although the technology itself was a well-established technique for assembling parts at high speeds, it had never been considered for the product in question.

One reason was that the surface states of the component materials were tacky. This tackiness created drag in the bowl feeders and tracks, which limited the speed at which the components could be fed.

A second reason was the necessity of deforming one of the components to permit it to "roll over" the other component. This rather tricky operation had always been carried out via pressurized air, with the part stationary. The time for performing the operation was not compatible with high-speed assembly.

Given these technical uncertainties, the initial investment was limited to an R&D study designed to develop new methods of feeding the components and to prototype an alternate means of carrying out the rollover operation. Fortunately, some solutions were found and validated, generating the confidence needed to pursue development of the ensuing machine. The initial investment in the study was only 5% of the machine price. Nevertheless, the resulting system performed at eight times the rate of the existing machines, with scrap reduced from 5% to 0.5%.

A word of caution is necessary with regard to prototyping—prototyping everything that is unknown or fuzzy can be impossible or prohibitively expensive. In some cases, the cost can be equivalent to making the machine or one of its subsystems itself. Thus, a balance must be reached so that the preliminary phase of R&D and prototyping is enough to justify moving forward with confidence, but not so extensive as to delay unnecessarily the project's launch or to skyrocket the customer's investment. This can be achieved, provided the customer agrees to accept the risk that cost overruns could occur during the debugging phase and that responsibility must be taken for those that arise from issues that were impossible or uneconomical to address during the preliminary engineering phase.

A second example from the dispensing systems industry illustrates this point. For years, trigger sprayers have been produced on intermittent-motion machines. But two years ago, a major company initiated a massive project to build a machine that would more than triple the standard industry production rate. The challenges, as well as the risks, were enormous.

Achieving the targeted speed was only conceivable by employing continuous-motion technology. Yet this technology had historically been limited to assemblies that featured round, symmetrical component parts. These "friendly" contours do not exist in the nonround, irregularly shaped components of a trigger sprayer. Earlier, research and development in the continuous-motion assembly of nonround parts persuaded the customer that such a project was worth pursuing. Yet the project's scale was found to necessitate further studies and prototyping.

These studies were carried out, but were limited for reasons of time and investment. Although sufficient confidence existed at their completion to proceed with detailed machine engineering, there was enough uncertainty to necessitate the maintenance of a price envelope. It was not until the completion of engineering that a firm price was quoted for the system.

Despite these precautions, part-design changes during machine construction, combined with the unexpected and unpredictable behavior of some of the components, revealed further challenges and cost overruns. The customer has recognized that the additional amounts to be paid at this point would have been far costlier in terms of both money and time had it commissioned exhaustive prototyping before embarking on machine construction. Risk has not been eliminated, but a balance struck to permit it to be appropriately managed.

Such instances demonstrate that a phase of preliminary engineering can offer enormous value. The builder can now afford to assign engineering resources to the project without hedging against lost quoting hours. The technical challenges of the project can be seriously examined, with tests conducted and prototypes constructed as needed. If changes to the product design can simplify its assembly, these changes can be identified and evaluated jointly. Key suppliers can be consulted and firm quotations for subsystems received. The price envelope can be eliminated, or at least dramatically narrowed. In short, the risk level drops and the confidence level increases proportionately.


If the preliminary engineering study demonstrates feasibility and validates the original concepts, it should also yield a firm and optimized machine price and a detailed description of the functions and features of the proposed machine. The price is optimized in the sense that any fudge factors thrown into the price calculation to offset unknowns have been eliminated or drastically reduced. Both machine builder and customer should be confident in the project's viability and therefore be ready to proceed with detailed engineering and construction of a successful machine.

On the other hand, if R&D proves fruitless or it reveals either a faulty concept or that the machine cannot be built within the estimated price envelope, appropriate decisions can be made and the project can be reconceived or abandoned. Even in these worst-case scenarios, the loss to each party is minimized, because neither has made a major commitment of time or money.

Here one might object that in establishing such a partnership, the customer is put at a disadvantage by making a premature commitment to a given vendor. After all, would it not be better to proceed in traditional fashion, collecting estimates from a range of vendors in order to establish a general sense for the price, lead time, and feasibility of the project, and then request a second round of firmer prices once things become clearer? Is it not better to keep everyone honest, delaying the awarding of the contract until the last possible moment in order to ensure that the final price is optimized through competition? Such tactics may work when the project involves a proven or standard machine concept or technology. Projects requiring innovative solutions, however, require more than adapting an existing production process for a new application. Rather, the machine must do something that has not been done before.

Perhaps the goal is a machine that runs significantly faster than existing machines for similar products, or one that reduces scrap by an order of magnitude. Perhaps an innovative technique is envisaged for carrying out an operation, a technique that has never before been implemented. Perhaps it is theorized that by pretreating a component in a certain way (e.g., heating it, cooling it, texturizing its surface, etc.), it will exhibit behavior more conducive to assembly. Whatever the particular innovation contemplated, the fact is that both customer and supplier are taking a step into the unknown. In such situations, risk cannot be avoided, but only hidden or postponed to the customer's disadvantage.


The machine builder may raise an objection as well. Most builders are vertically integrated, which means they rely on their engineering department to generate downstream work for their other production departments, such as machining and assembly. R&D projects tie up engineering resources without guaranteeing any downstream benefit to the rest of the company. Thus, the machine builder must accept an opportunity cost of engaging in R&D, risking that an irreparable hole will occur in the production departments' schedule if the preliminary study does not lead to a machine construction order.

This is only a problem if the machine builder has no independent commitment to research and development. If the builder is not dedicated to expanding the range of applications for the firm's technology or to staking new claims in uncharted territory, then the opportunity cost of engaging in R&D may indeed be too high. On the other hand, if the builder is committed to innovation, then this opportunity cost will not be viewed as a cost at all, but rather as windfall financing.


In summary, innovative automation projects require identification of the risks that threaten both customer and supplier, an agreement on a fair method of sharing them, and completion of one or more steps of limited preliminary engineering analysis before a firm and mutual commitment is made to the machine. In this manner, risk can be truly managed to the benefit of both parties, rather than hidden or ignored to their detriment.

Lawrence Allingham is vice president of Lagniel Inc. (Wood Dale, IL). François Paillusseau is CEO of Lagniel S.A (Douvres-la-Délivrande, France).

Copyright ©2002 Medical Device & Diagnostic Industry

500 characters remaining