Excellent management of talent, domain expertise and good processes are all essential. But to truly maximize productivity and return on investment, NPI organizations need to not only run efficiently, but also have strategies that will get the most out of prior R&D investments and take advantage of the available technology ecosystem. This means leveraging hardware and software re-use as well as other resources available from the industry. Re-inventing the wheel can too often bring engineering organizations to their knees. It is important to have a strategy for re-use and to carefully decide what is of strategic importance to develop in-house versus what can be outsourced or obtained from third parties.
While the topic of re-use is abstract enough to make a formulaic approach impossible, there are several guidelines that can significantly improve NPI results. Starting with hardware, here are a few examples:
- Common chassis can be used for multiple solutions with minor modifications across products
- Modular design methods can allow broad re-use of switch, CPU, I/O, and power supply modules
- Entire platforms can be re-used with different software to distinguish each system
- Because complex hardware components or sub-systems require significant software support, component re-use significantly accelerates development and test cycles (such re-use also significantly improves buying power)
With a focus on hardware modularity and sub-system and component re-use, new hardware development cycles can be dramatically accelerated at much lower cost.
From a software perspective, there are other similar actions to take. Namely:
- Converge system operating systems to avoid duplication of effort across different products
- Eliminate multiple copies of the same functions and re-use common software components
- Develop Object Oriented software to provide interfaces that are readily reused and shared
- Re-use entire software implementations with features used or disabled depending on the product solution, price and target market (subsets and supersets of functionality)
- Design in software licensing support to allow software functionality that depends on configuration and license key installation, allowing one software solution to be sold as several solutions
- Intelligently re-use open source software, especially where functionality is not tied to any specific competitive advantage (more on this later)
By implementing a strategy for both hardware and software re-use, most organizations will see dramatic improvements in time to market, portfolio breadth and quality. (Quality is improved because you only have to debug the systems once!) If your team does not support a re-use strategy, you definitely need to understand why.
The next productivity opportunity to discuss is how you should use third parties in development. While there are many ways to leverage third parties, I will highlight just a few here from both hardware and software perspectives:
- An ODM model (original design manufacturer) can be used to augment your engineering team. You provide design specifications to a third party with specific domain expertise who implements the design for you, for a fee, or manufacturing rights, etc.
- Industry standard chassis or platforms (such as ATCA platforms, or Intel appliance servers) speed time to market and reduce development cost (but may also reduce margins!). You can also borrow from industry standards to reduce the cost of from-scratch development speed time to market.
- Hardware reference designs from silicon providers (CPU, I/O, switching silicon, etc.) can be used to minimize risk and accelerate time to market.
- Open source software should be used whenever high quality code can be found. Because it is readily available to your competitors, you are at a disadvantage not to use it too. (However, you must fully understand HOW to use it so that you do not sacrifice your proprietary intellectual property to the open source community!)
- You can also use licensed software as appropriate. In many cases, it is simply not feasible to implement all of the software you need from scratch and many specialized and high performance functions may not be available form open source. When licensing software, however, it is critical that you properly negotiate reasonable fees as mistakes here can be very damaging in the long term.
In all of these cases, you should endeavor to leverage partners as much as possible, especially when it costs you nothing more. For example, in using reference designs from a silicon provider (a CPU, for example), you should only award business if you also get assurances of support, such as design reviews, integration assistance, etc. In negotiations with suppliers, all things are possible and you should avail yourself of every advantage that you can.
Ultimately, what you want is a “Hybrid Model” for your development and R&D efforts. That is, you want to leverage third parties where they accelerate your efforts, where they do not compromise your differentiation or defensible differentiation investments and where they are affordable. You also want to leverage re-use in both hardware and software to minimize waste and speed time to market for multiple solutions. You want to put the lion’s share of your own R&D efforts into core technology that is at the heart of your differentiation, spending your most talented resources on the pieces of technology or specialization that make your products and services unique.