RFID Research Supports Real-World Experimentation

By Kurt Hozak

Simulation can be a good tool for modeling the effects of known potential RFID process changes, but real-world experimentation may uncover additional benefits and changes.

In a recent article, RFID Journal founder and editor Mark Roberti suggested that businesses should consider thinking of radio frequency identification as infrastructure, since it can make such a wide impact within organizations and across supply chains (see Do You Really Need to Justify Your RFID Investment?). When RFID is broadly used in this way, he argued, traditional return-on-investment (ROI) justifications focused on specific applications become less critical. This is important, because even with falling costs for tags and supporting hardware, some companies remain hesitant to make RFID investments when they focus on individual applications and expected benefits. To help overcome this concern, Roberti and others have encouraged businesses for years to experiment with RFID, even if only on a relatively small scale, because surprising benefits can often be achieved by using the technology (see Mastering the Tool and New Balance Taking First Steps With RFID).

Academic researchers recently reached similar conclusions about the shortcomings of traditional payback models that assume too few benefits from RFID (see An RFID Application in Large Job Shop Remanufacturing Operations). Professors Geraldo Ferrer, Susan K. Heath and Nicholas Dew, from the Naval Postgraduate School (NPS), examined the use of RFID as part of a real-time location system (RTLS) implemented to help improve remanufacturing performance at the Tobyhanna Army Depot, in Pennsylvania. By implementing RFID, the depot was able to reduce the average time required to remanufacture one type of radar system by 62 percent, though remanufacturing processes for other types of radar systems showed less improvement. Because of this, the researchers developed a simulation model using the Arena software package to identify process characteristics that facilitate larger performance gains.

In order to obtain more generalizable results, the simulation was not intended to be an exact representation of the various processes undertaken at Tobyhanna. It initially focused on a few ways in which RFID can change remanufacturing processes (by facilitating faster parts sorting, for instance). Although it was anticipated that the studied changes would have a large impact on overall performance, the simulation results showed far less benefit from using RFID than what was being achieved in the high-performing real-world process. The researchers realized that the initial scenarios being modeled did not include all potential process changes made possible by using RFID, so they re-ran the simulation with a new scenario that included a broader range of process improvements (for example, better prioritization of resources as a result of improved information). The new scenario showed overall flow-time reductions closer to what was actually being achieved by the high-performing process. Upon meeting with the manager of a remanufacturing operation that did not show much improvement, the research team observed that the process had not made the same changes to fully take advantage of RFID, which was the situation modeled during the initial scenarios.

If the Tobyhanna Army Depot had conducted an ROI analysis prior to implementing RFID, and had only assumed that a narrow range of process changes were possible, it might not have made the investment that led to the high-performing process. The research shows that analysts should not be overly conservative when estimating RFID's cumulative impact. The NPS researchers have very strong backgrounds in industry and academia, so they realized that the initial simulation scenarios that they studied had not reflected all ways in which RFID improves processes. Although they adjusted their analysis accordingly, it is not difficult to imagine others employing a less rich model and erroneously concluding that RFID is not worth deploying. After the technology occasionally experienced excessive hype over the years, it is ironic that companies may now miss out on significant competitive benefits due to a lack of imagination and an excess of caution.

The comparison of the Tobyhanna processes and the corresponding simulation scenarios provides further evidence that simply implementing RFID without enacting process changes may not be sufficient to demonstrate large benefits. Research that I conducted with David A. Collier (see RFID as an Enabler of Improved Manufacturing Performance) showed that if RFID is simply used as a bar-code replacement without actually taking advantage of it to enable new or improved processes, the technology might offer little benefit. I later noted in follow-up research conducted with James A. Hill (see The Material Handling Trade-Off When Lot Splitting With RFID in a Job Shop) that because additional investments may be required to support those other changes, "managers should therefore not think of RFID as an investment to be implemented in isolation... but as a technology investment that both enables and requires new operating policies and capabilities to support overall strategic objectives." Congruent with those research findings, and with Roberti's suggestion to think of RFID as infrastructure, the NPS research team asserted that "...RFID technology should be conceptualized and modeled as part of an 'innovation bundle' rather than as a standalone technology adoption."

Although the NPS researchers noted the difficulty in fully capturing RFID's holistic benefits, simulation offers several advantages compared with other methods of modeling the RFID "innovation bundle." Simulation has long been regarded as a relatively flexible and powerful tool to help users understand complex relationships that can not be easily reduced to equations. Simulation's ability to visually model processes can help analysts more quickly develop models that are richer than spreadsheet-based ROI calculators, and managers may have more confidence in the results if they can see the source of the numbers from the simulation. By contrast, parameter estimates or output in spreadsheets may be more difficult to accept if one is unable to see justification or evidence for their values as clearly as with simulated processes.

In summary, companies may need to make a variety of process changes in order to achieve significant benefits from RFID, similar to other infrastructure investments. When used appropriately, simulation can be a good tool for modeling the effects of known potential RFID process changes—but real-world experimentation may uncover additional valuable process changes unforeseen, or previously believed to be relatively unimportant. Even if the process changes are known, incorporating them all into traditional ROI models may be difficult, which may lead to underestimating RFID's cumulative benefits. The NPS professors should be commended for producing research offering such practical relevance.

Kurt Hozak is an assistant professor of operations management at Coastal Carolina University's E. Craig Wall Sr. College of Business Administration, and a technology and operations management consultant.