Viewpoint News, January 2010

Resource Limitations

by James Campbell - jac@viewpointusa.com

In the past three months, I have discussed topics from the ‘7 Challenges for Highly Effective Test Engineers’ that are fairly technical. This month’s topic starts the discussion about steps actually needed to build or upgrade the test system. Specifically, I will focus on the under-appreciated topic of resource limitations. The scenario starts with you being given a project with an unbelievably short timeline, using unfamiliar technology, and you are not sure how to complete the project on time.

No Way Will This Be Done in Time!

I can think of three main resource issues that prevent projects from reaching milestones on time and in budget. These issues are:

• Too many new measurement and control requirements to satisfy in the allotted time.
• Unfamiliar expertise needed to satisfy the requirements that may take too long to learn.
• Accessibility of hardware required for implementation.

Of course, there are other issues, but, in my experience, these three are the 80% of the usual 80-20 rule. Let’s address each of these individually.

Too Many New Requirements

Try to reduce the burden of new requirements. Is it possible to introduce some requirements in a phased delivery approach? Are some requirements less important than others and perhaps not actually required? Keep only the most important if it will help you meet timelines.

Try to reduce the time to implement the requirements. Can the requirements be combined or can execution sequence be reordered to minimize the development time? Look for existing test code that satisfies the requirement with a superset of functionality; perhaps it does more than needed, but since it’s already developed, you can eliminate effort.

Unfamiliar Expertise

Spend a few hours estimating the effort to learn the new skills or knowledge. The classic saying goes, “You don’t know what you don’t know.” How familiar are you with the FPGA board that this project is requiring you to use? Does a driver exist to interface with your new oscilloscope, or do I have to invent one from scratch? Has anyone in your company already developed an architecture to accomplish what you are trying to do?

By sitting down and attempting to answer these questions at the onset of a resource restrained project will give you perspective into the effort that is going to be needed to implement all of these new features. Given this knowledge one may be able to influence what can and cannot be done in the time line to deliver a successful project. As corny as it sounds, knowledge is power, spend some time educating yourself on the technology you are going to need for your project.

Meet with the product vendor. Does the expertise revolve around a new instrument or measurement technique? You might get enough training from a vendor or distributor that supplies the instrument or sensor. Sometimes, if you needs are simple enough, they might just do the work for you.

Hardware Access

Identify hardware usage. You may find that the equipment you need for development is also needed in some other group, such as a design engineer, an integrator, or another test engineer. Obviously, this resource limitation will impede your progress if both need at the same time. Plan ahead and alert the other group(s) as to your expected usage.

Identify parallel path development efforts. Design and manufacturing development efforts are frequently done in parallel. Without an actual unit to validate your test system, you must simulate the unit under test (UUT). Unless you have strong confidence that the UUT will be available in time for system validation, I recommend you begin simulation efforts as soon as possible and be prepared for many changes, since the product is still in design.

Conclusion

Most resource limitations can be itemized near the start of a project, and especially by the end of the design phase. Proper planning is the key. If you’ve not already done so, interview for outside capability before you are in a crunch. Outside resources can assist or consult on software architecture and development, UUT simulations, selecting appropriate measurement hardware, or fabricate automation equipment.