05-30-2008, 10:28 AM
|
|
Your Unique, Evolving RFID Plan (Part 1)
Issue #148 | May 30, 2008 | by Venkat Krishnamurthy
If you are reading this column, you may already have a good idea of what process or processes you’d like to RFID-enable, if you are not already doing so.
You are also doing your homework in advance; that’s good. It means you are interested in better understanding the critical issues to be considered before tackling your RFID initiative. That’s important, since the approach taken can easily mean the difference between success and failure.
Let’s start with a basic fact: No two RFID deployments are exactly the same. I say this, because most of what follows is based upon that reality. That means that your implementation will almost certainly be different from a comparable project underway at similar company -- and is likely to evolve over time.
Here then are some recommendations:
Match your business environment to RFID processes: The best way to do this is to use pre-packaged workflows wherever possible, taking them one step at a time. From there, you’ll likely customize them to meet your specific needs.
Avoid building workflows from scratch – there’s really no need for that. In the unlikely event that you find a solution that meets your needs out of the box, by all means take advantage of it. But more likely, you’ll find workflows that meet most of your needs and will then tailor them to complete the fit.
Future-proof your RFID deployment: Remember that your needs will evolve. So ensure that you are using easily modified code and that it is well documented. Every RFID deployment I’ve been associated with has started small and grown over time.
Human nature is such that people realize benefits they hadn’t thought of earlier, or figure out new ways to increase efficiency. So avoid shortcuts that leave you with millions of lines of code that aren’t documented. Good libraries and well-documented code will make your life (and your organization’s!) much easier.
Build the Foundation for Ongoing Development: Take into account that over time you are likely to use new devices, introduce new processes, add new locations, etc. For example, devices come in all kinds of modalities: GPS systems, bar codes, RFID readers, etc. Your RFID strategy should allow for the inclusion of a variety of these devices. Moreover, since pilots may grow into enterprise deployments, scalable and configurable strategies are critical. If you do not commit serious forethought to this, you’ll find yourself reworking your RFID solution for each individual site.
Make sure you also devise a workable data warehousing strategy that analyzes data coming out of the process. I recently worked with a manufacturing customer that implemented RFID on the factory floor. Based upon its analysis, the manufacturer discovered the mean and medium times of certain time consuming operations. As a result, the firm was able to shave 20% off of those times.
Finally, give some thought as to how you will keep your RFID system separate from, but still integrated with, back-end systems. By this I mean that you want to keep your RFID execution process independent of those back-end systems to avoid having data trapped in them – while keeping the systems integrated so data can be used for other purposes.
Next week, we will share some examples and thoughts on using these guidelines to select your development approach.
Venkat Krishnamurthy is chief technology officer for OATSystems, a provider of RFID solutions (www.oatsystems.com).
|
