When are seemingly more expensive mass production methods the way to go vs. prototype or low production? To answer that question, you'll need to follow some important steps.

First, you'll need a pre-clinical pilot clinical study, or possibly a beta launch to test the market. You don't want to spend more than needed, but you also don't want to paint yourself into a corner if you have to make more units than you anticipated.

When developing a product, a good practice is to use real production processes, real production methods, and real production materials wherever possible. And apply design for manufacturing principles as early as possible. “Hardware store” materials are good for conceptualizing a product and initial proof-of-concept bench testing. However, when progressing beyond this, think like a pool player. Take your shot, but set up your next shot in the process. This way you won't wind up “behind the 8 ball.”

Planning is especially critical when developing FDA Class III devices. Once the PMA dossier is set, re-opening with material and design changes can be very difficult and expensive.

Rapid prototype options

RP technology is faster and more accurate than ever. Plus, material choices are better. Achieving the right combination of material, surface finish, and feature resolution remains challenging. Methods that can produce small feature sizes, such as Polyjet 3d printing technology by Objet Geometries, Ltd, Billerica, MA, (objet.com) are useful for prototyping close-tolerance fit check parts and even some larger catheter tubes. It is also capable of rapid prototyping rubber parts. Medical grade polymer material options are available via the 3D System by Stratasys, Eden Prairie, MN (stratsys.com). Direct-metal e-beam and laser-sintering systems by Germany-based EOS (www.eos/info) and Sweden-based Arcam (www.arcam.com), now make possible titanium and cobalt-chrome implants possible without expensive investment casting. These and other RP methods are capable of producing devices such as orthopedic implants and hearing-aid shells.

It is possible to use RP to make shapes and structures that are difficult or impossible to duplicate with conventional machining and molding. However, this capability can have negative consequences if the prototype is made with unmoldable features such as undercuts and lack of draft angles that require redesign for molding later.

The best place to go to see the range of RP solutions available is the annual RAPID show, sponsored by the Society of Manufacturing. Next year's event will be in Anaheim, CA, May 18-20 (www.sme.org.rapid).

Short-run manufacturing options

The tradeoff here is usually between lower setup and tooling costs and higher per-part costs. Remember, it's the total costs and the ability to produce a final usable part in the fewest number of steps that count. At the lower end there is vacuum forming, casting in RTV (room temperature vulcanizing) silicone molds, machining, pressure forming, foam molding and RIM (reaction injection molding). Typically these methods are most useful for enclosures. For devices, machining may be a good option for metal parts and smaller quantities of items like plastic handles. If you are going to make larger quantities later, check that your material is appropriate and available for production. You don't want to validate a part in a machined material that is not available or inappropriate in an injection-molding process. You may also avoid the cost and having to repeat expensive biocompatibility testing. For medical devices extrusion can be a useful option, as the tooling costs are fairly low compared to injection molding. Again, design for manufacturing early, and plan your shots.

Short-run injection molding

Injection molding in aluminum tooling has become more readily available, cheaper, and easier to use. An advantage of short-run injection molding is the ability to mold parts in the final production material, or the ability to use materials not available in stock shapes for machining. Typically, parts for medical devices are small and therefore the shot sizes are small. Many resin manufacturers are willing to send a sample of resin (about 20 lbs) for evaluation. This may be enough to run hundreds of smaller parts. This way you have parts that are in their production form, rather than a facsimile version, for bioburden, package validation, and biocompatibility testing. Short-run soft tooling can also serve as bridge tooling to produce parts for clinical trials and marketing samples while production hard tools are being made. One company that specializes in aluminum soft tooling is Protomold, Maple Plain, MN (protomold.com). CAD models can be uploaded from its website and checked for moldability (e.g. wall thickness, draft angles, and undercuts). A cost quote for tooling and piece parts may also be generated. Protomold has in-stock resins or it can mold customer-supplied material.

Production tooling

When you get to this level, the last thing you want to make is an expensive, finely machined Class A polished “boat anchor” (useless mold). Take the necessary steps to be sure that what you can make lots of are 1) what you really want, 2) really work, and 3) will really sell. One way to save money at this stage is to use “MUD Base” tooling (Master Unit Die). This is a modular mold frame system that may be cheaper than a fully custom mold. Use less-expensive tooling, to start, and take the money you make on the product to pay for better tooling if the product is a winner. If it's not a winner, you have avoided buying an expensive “boat anchor.”

Another powerful tool for controlling overall costs is to design out labor costs whenever possible and make the mold do the work. Labor costs and secondary operations can be difficult to remove once they are “baked into the cake.”

A simple decision tool

It's the total costs of your development program that count. The goal is to get where you need to go in the fewest number of steps, while spending the least amount of money.

The graph on page 30 (Fig. 1) you can run in Excel or MathCad. Put your setup costs on the Y axis, and plot the cumulative costs (setup + part costs) slope on the X-axis. You may be surprised to find that in the quantities you'll need to make, a higher production method with higher tooling and setup costs, is cheaper overall than a lower production method with high per-part costs, high labor content and no economies of scale. Reducing the data to a chart can help you decide and can help sell the program to management as well.

Knowledge pays

Knowing the range of options to make the best design decisions will help you get to your end result in the fewest number of steps, spending the least money. Remember, the method with the lower startup costs may not be the cheapest in the long run when cumulative costs and tradeoffs like biocompatibility of materials are taken into account.