Section 9.3. Implementation Notes

9.3. Implementation Notes

This section makes several suggestions based on experience gathered from implementing real-world RFID systems. The suggestions are free-form and are listed here without any order of priority:

• Test, test, and test.

• Analysis first, hands-on next.

• Avoid scope creep.

• Need for multiple tag checkpoints.

• Composition relationships are important.

• Never overlook tag validation.

• Include expectation management.

• Remote monitoring and management capabilities are essential.

• Time process changes.

• Comply with regulations.

• Need for patience and hard work.

• Unexpected benefits.

• Involve smart people.

The following subsections explain what these mean.

9.3.1. Test, Test, and Test

An RFID system should be tested as thoroughly as possible, and then some. This is especially true in an iterative approach where the current system release has to be as bug-free as possible so that the next iteration can be built on top of it. Tag placements, antenna positions, and so forth all have to be tested rigorously until the system performs at a satisfactory level at its current scope. At that point, the system can be moved to the next iteration. Several vendors and associations have set up RFID laboratories that you can use for testing purposes. These labs lack real-world context, but provide opportunity to test tags without disrupting business operations.

9.3.2. Analysis First, Hands-On Next

It might come as a surprise to you how much information can be gathered by upfront analysis instead of blindly diving into the hands-on exercise instead. The important process flows and use cases should be analyzed as much as possible to glean the business characteristics and parameters before any hands-on activity begins. After the designers are confident that they have grasped the essentials, setting directions and strategy follows naturally, making the hands-on trials and projects much more focused and useful. Without this guidance, hands-on activities might dither and prolong unnecessarily. In one actual case, a business spent more than a year of hands-on experimentation without really coming to a conclusion to use (or not use) RFID!

9.3.3. Avoid Scope Creep

Laser-like focus should be maintained on the scope of an RFID system. It should not diverge to include features that were not originally included when the design and analysis work was done. It is easy to get carried away by the success of a throwaway pilot and be unrealistically optimistic about the technology prospects. For example, the scope might include changing business processes when the other parts of the system are poorly understood or analyzed. As a result, the system's scope, scale, and complexity might grow to a point that the analysis, design, and implementation encounters major roadblocks that jeopardize the delivery timeline. Scope should be handled in iterations. Scopes should be prioritized using some criteria (such as ROI, risk, and so on) and then iteratively implemented in such a manner that the total risk is always controllable for each such iteration.

9.3.4. Need for Multiple Tag Checkpoints

A reader validates a tag at the point of its creation. However, additional checkpoints are necessary, either manual or automatic, at other points in the operation. The reason is simple. A perfectly working tag might become damaged after it passes through a certain number of processing steps in the operations, rendering it completely useless. If a read of this tag is then attempted at any point, the read will fail. To catch this problem, multiple checks needs to be put in place in the operations to make sure that a tag is not damaged and can be properly read. If a damaged tag is found, corrective actions need to be taken to fix the problem (for example, creating a new tag with the unique identifier of the damaged tag and attaching it to the item). As you can understand from this discussion, merely slapping a tag on an item is no guarantee that it can be read at a later point in the operations. Therefore, even a slap-and-ship type application needs to validate tags after they are created and put on an item, most probably when items are leaving a business site (for example, when the item is being loaded onto a delivery truck for its customer).

9.3.5. Composition Relationships Are Important

The interrelationships among entities such as pallet, case, and unit are important. Therefore, a tag ID of a pallet (when loaded) should correspond to a set of case tags that this pallet contains, and each such case tag should then correspond to a set of unit tags that this case contains. Although unit-level tagging is not prevalent today, these are some examples of fundamental composition relationships that exist in business operations. Other composition relationships might also exist in a business. An RFID system should maintain these relationships instead of being "flat," where the associated tag data for a pallet, case, and a unit are implemented at the same level of hierarchy.

9.3.6. Never Overlook Tag Validation

It is easy to get hung up on tag readability and overlook an equally important aspecttag validation. Tag validation means that the tag data is checked against some criteria to make sure that it is valid. This is generally done using business-specific logic and data stored on the back end in conjunction with the tag data. A tag whose readability is close to perfect but contains data that is invalid is not useful to a business. You would be surprised how often this attribute is sacrificed in favor of tag readability.

9.3.7. Include Expectation Management

Expectations need to be managed carefully in an RFID system implementation effort. The designers and implementers need to make the stakeholders aware of the benefits that can be realistically achieved from such a system. Business needs to be patient about expecting savings from a newly minted RFID system because this system might need time for fine-tuning to perform optimally. A slew of business practice eliminations or changes should not be undertaken with the assumption that an RFID application will replace or provide a superior alternative to these practices before actually validating that the application can deliver on its promises.

9.3.8. Remote Monitoring and Management Capabilities Are Essential

Remote monitoring does not seem to be a necessity until some system component fails and then business operation is impacted to troubleshoot and fix the failure. An RFID system might not be maintainable unless a monitoring capability is built in to the system. Similarly, if a business site has 200 installed readers, manually upgrading the reader firmware to its latest version will need a substantial amount of personnel time and business downtime. To make matters worse, some of the readers might not need the upgrade or might need a different version of the firmware, thus making any manual update procedure prone to error. Remote management is an effective answer to this type of issue. Therefore, designers and implementers need to pay special attention to include these features in the system, which will make the system efficient and economical in the long run.

9.3.9. Time Process Changes

Although it is true that the maximum potential of RFID can be realized through the existing business process changes, it is truer that these changes should not be introduced at the early stages of implementation. In other words, unless the RFID system has proved itself in the actual business operations, the business process changes should not be implemented. Why? Because if an RFID system does not deliver according to the business expectations or needs time for fine-tuning, it can be isolated without causing any impact on the existing business operations. After the RFID system has matured to a certain acceptable point, the business process changes can slowly be introduced.

9.3.10. Comply with Regulations

Introduction of RFID systems should not violate any federal or state regulations or job safety codes. The business can be subject to hefty fines and might have to assume substantial liability if it does not comply with such regulations. For example, some items (such as pharmaceuticals) might require you to follow strict government guidelines for tagging; the hardware installations might need to be done by properly trained professionals who are knowledgeable about these codes and regulations (which might also reduce hardware damage and system downtime).

9.3.11. Need for Patience and Hard Work

No single one-size-fits-all RFID solution exists. Likewise, you cannot just buy a plug-and-play solution and use it without any modifications. The situation is made more challenging by the fact that a similar solution that has worked for one business might perform badly for another. This means that implementing a successful RFID solution is not easy. The variables involved (as described throughout this chapter) need to be methodically analyzed and resolved to arrive at a system that satisfactorily meets its requirements. Therefore, patience and hard work are essential ingredients to successfully take on an RFID project.

9.3.12. Unexpected Benefits

A pilot or an evaluation effort might reveal unexpected advantages that were not comprehended at the planning or design time. Such benefits might manifest themselves in the form of automating parts of the business as a side effect of the scope of the pilot, substantially more productivity gains than originally expected, and so forth. Therefore, implementing a pilot can offer additional advantages besides technology evaluation and validation of business assumptions.

9.3.13. Involve Smart People

Designing and implementing a nontrivial and successful RFID application takes a lot of effort by bright people. If a business expects to engage second- and lower-rate talents just because these individuals are available, be forewarned. The early stages of RFID adoption should be in the hands of the people who not only have deep technical skills but also business domain knowledge. These people should be able to separate reality from hype, set realistic expectations with the stakeholders, and have unrelenting drive to apply and improve the system to exploit its optimum potential.