To buy or not to buy: factors influencing whether to do-it-yourself, or use an off-the-peg solution.

In the military equipment world, where cost over-runs and delays are all too commonplace, there has been a strong push towards using 'Components Off The Shelf' (COTS), as a way of reducing design cycle times and cutting manufacturing costs - COTS often reduces COST. The arguments which hold good for military contractors also apply in many other areas such as consumer goods, medical equipment and transportation etc. Deciding on a way forward An OEM will always want to personalise his products externally via styling or unique features for instance, but the inside can contain bought-in components such as printed circuit board assemblies, power supplies and displays without comprising on the unique look and feel of the product. In a complex product such as a mobile phone, the uniqueness or distinguishing features of a product will often be decided largely by the application software, which is what the user sees and interacts with. But this application software is only one of many layers , and the lower levels of software can also be regarded as COTS components. The COTS Philosophy The COTS philosophy has been successfully applied in not only the hardware - PCBs (Printed Circuit Boards) , enclosures, PSUs ( Power Supply Units), but also the software - operating systems, drivers, GUIs and stacks to name just a few. COTS solutions are available at different levels of integration, including • Complete boxed solution (with enclosure, display, PCB, PSU etc.) • Single-board PCB • Module (SOM)+ baseboard PCB All offer the potential to shorten development time which can deliver reduced design cost, faster time-to-market and quicker ROI. These factors combine to reduce design risk, accelerate software development, prevent obsolescence, improve upgradeability and in some cases, second source availability. Engineering vs. Business The factors which affect the make-vs-buy decision fall into two main categories: technical and business-oriented. The technical factors can include performance, legacy support and backward compatibility, conformance to standards, form factor, power consumption and heat dissipation, manufacturability and testability. Some of the business-oriented factors include: • Time pressures • Development costs • Production costs • Risk • Production quantity/ sales forecasts • Availability of in-house expertise and resources We will be examining these factors in detail later in this article, but for now it is sufficient to say that every situation is different, and every customer will have his own priorities. Our experience as a supplier of modules and other off-the-shelf solutions, and also as a custom design-house, gives us a good vantage point from which to assess a customer's needs and to make appropriate recommendations. One thing we have learned over the years is that in most cases, the volume threshold for a standard or COTS module is surprisingly high. This is mainly because many OEMs do not place a realistic value on the engineering effort involved in designing a complex module, and setting up the appropriate systems to support it throughout its life. It is all too easy to underestimate the cost of bringing a new product to fruition, and maintaining it thereafter - and that's just on the hardware side. Software as a commodity The introduction of software into the mix brings a whole range of additional factors to be considered, and there are many examples of software not being considered at all in the early make-vs-buy discussions. ("It's only software" or "its only noughts and ones" being common mantras). With the availability of off-the-shelf BSPs and middleware components such as GUIs, file systems and communication stacks, the software designers' job can be significantly simplified, enabling him to focus on his application, and not on why the darned thing won't boot! The full article explores: • The cost implications of make versus buy • Costs and compromises • COTS computers • Choosing the right form factor • Real-life case studies where the make vs. buy argument was applied