Reaching the Decisive Moment

By John Edwards

Transforming an RFID pilot project into a real-world deployment demands vision, attention to detail and a determination to meet unexpected challenges.

"The decisive moment," a term coined by the pioneer photojournalist Henri Cartier-Bresson, refers to the point at which all the fleeting elements necessary to create an awesome image converge in the camera's viewfinder, signaling the photographer that now is the time to press the shutter button.

Organizations that have deployed RFID pilots face their own decisive moment. It occurs weeks or months after the trial winds down, when precise planning, hard-won experience, test data analysis and business insight all merge inside the project leader's head, telling him or her that now is the time to roll out the system.

Illustration: iStockphoto

It all sounds so simple—but decisive moments aren't all that easy to spot. Cartier-Bresson, who shot stunning photographs worldwide for more than half a century, admitted that it took a great deal of experience to know exactly when to push the shutter button.

RFID project leaders face a greater challenge in determining when to push a pilot into final deployment. "You always wonder, 'Is this a good time? Is this not a good time?'" says Philomena Sousa, oncology systems manager and process specialist at Toronto's Odette Cancer Center, who worked on the team that created an RFID-based patient arrival and tracking module. There's always a feeling, she says, that there's another improvement to be made, another issue to consider.

Rolling out a pilot isn't simple, and specific issues vary among companies. Yet, there are common situations and challenges that nearly every RFID project leader will face as a system is being prepared for full operation. Project leaders who address these issues can feel more confident that they'll have a successful launch and the solution will deliver benefits for the long term.

Start With a Plan
The key to a successful rollout begins with meticulous planning at the pilot stage. In RFID Journal's story Eight Common Deployment Mistakes—and How to Avoid Them, RFID experts cite the importance of selecting an experienced systems integrator, choosing the right RFID technology, assigning responsibility to one senior executive, thinking strategically and handling the integration of RFID data with back-end systems.

"I'm a big believer in pilots, but I also believe that there's a lot of work to be done before the pilot," says Steve Halliday, president of High Tech Aid, an RFID consulting firm based in Gibsonia, Pa. "I know some companies would rather just throw together a pilot and then work on solving the issues that come up when it's time to deploy, but that doesn't work for me in my company," he says. "We like to try and identify as many of the issues before we start as we can."

A pilot heading out in the wrong direction will leave a business ill-prepared for a successful rollout, says Grant Richardson, director of the innovative solutions group at CDO Technologies, a Dayton, Ohio-based RFID systems integrator. "First and foremost, you need to understand what problem you're trying to solve," he says. "There definitely needs to be some sort of statement or vision of the problem." Setting goals is the next step, Richardson says. "What is your end state? Where do you want to be when this project is done?"

Steering the pilot in the right direction for deployment also requires establishing specific rules and benchmarks, says Bill Spahr, VP of professional services at Ekahau, an RFID provider based in Reston, Va. Project leaders need to ask themselves several basic questions before launching a pilot, he says, including: "What's the evaluation timeframe? What are the next steps? How do we evaluate success?"

Project leaders should carefully document their current processes before attempting to automate them with RFID, advises Ryan Mabry, a lead software engineer working on the development of an RFID-driven shipboard inventory system at the Naval Surface Warfare Center (NSWC) in Panama City, Fla. (see Naval Surface Warfare Center Demos RFID Tool-Tracking Solution for Combat Ships). This information, he says, is essential for developing a system that will deploy successfully. "A lot of times, people don't have the [current] process actually documented," he explains. "They don't know what they are doing today, or how they actually conduct an inventory, or even what they are actually inventorying." Key details, he says, are often "the hardest things to nail down" before beginning work on an RFID-based system.

Agree on the Results
As the pilot wraps up, project leaders, systems integrators and stakeholders need to reach a consensus on the trial's results, to determine whether it's time to move on to a full-scale deployment. "It's important to agree on success benchmarks up front," Halliday says. "If you've agreed on the right things, it should be very easy to say whether or not you've done what was needed to move forward." The problem, he explains, is that people sometimes don't know what they really want. "Then," he says, "it becomes much harder to agree up front what the measurement point is."

"Sometimes it's a no-brainer whether to flick that [deployment] switch quickly or slowly, but it's really... measured on a case-by-case basis," says Philip Lassner, CEO of Overheer Systems, a Westmount, Que.-based systems integrator serving retail customers. He recalls one project when a rapid rollout was expected but ultimately deferred. "We thought we were going to do this quickly," he says. Then the client requested a delay, sensing that things just weren't quite right. "It turned out there were some IT issues that needed to be addressed first," he says.

Businesses must examine their current business processes to define pilot acceptance criteria, Spahr says. "Ultimately, what moves the needle is pilot success against acceptance criteria," he observes. "In health care, for example, they would look at things like utilization rates, equipment cycle testing, and equipment being in use versus not in use."

Obtain Buy-In
It's important to ensure that top management remains committed to the project and will fully support—and fund—the system as it moves to the next stage. "Once you get that buy-in, everybody becomes a part of the success," Spahr says. "You want the system to be something organic that improves processes overall, and you only get that when you have full buy-in."

Buy-in represents management approval and support, not necessarily direct project participation, Halliday notes. "Top-level management doesn't always want to be involved, and there's a lot of work that can go on without them being involved," he says. "But I think it's really important to give them an understanding of what's going to happen and what it will mean to them."

Often, buy-in is achieved by demonstrating return on investment—that magic formula that can determine whether a pilot is worth taking to the next stage. "In today's economy, no one's got money to burn," Halliday says. "Businesses want to know up front if they will see a three-month, two-year, 10-year ROI."

The NSWC went to great pains to ensure that its RFID system would live up to its ROI goals, Mabry says. "We had to do a business case analysis, and that included return on investment, payback period, reduced workload, and what it would take to actually put all of this into implementation."

Many enterprises conduct time and motion studies and other sophisticated process evaluations before and during the pilot. But squeezing meaningful ROI insights out of such data prior to deployment can be tricky. "You can show that by doing things a certain way, you're going to shave three seconds off a step in a process," Halliday says. "Yet, typically, there's no hard and fast way to work it out, nothing that's going to be the way to compute this sort of thing."

Sousa convinced Odette Cancer Centre management to roll out its patient-management solution by showing various stakeholder committees the concept in operation. "Prior to going live, we had a number of demos," she says (see Odette Cancer Centre Boosts Throughput via RFID). "After the pilot, our findings were summarized and the analysis, the business case and the go-live plan were compiled and presented to the senior management for approval. After approval, the implementation team decided that the chemotherapy suite would be the first to be rolled out due to an immediate need."

In some cases, the "go/no-go" decision on a pilot set to deploy is complicated by the fact that ROI isn't simply a straightforward calculation based on time, dollars and equipment. Mitigating risk, or failing to mitigate risk, can have a significant impact on an RFID deployment's projected ROI. "Ultimately, ROI matters, but there are also a lot of intangibles, like safety," Spahr says. "It's hard to put that into a calculator. A single incident can be absolutely devastating to an organization, particularly in schools or in health care. Those things can be difficult to put a number around, but you certainly can't underestimate the value of them." (Deploying personnel safety solutions to locate employees during emergencies, for example, is typically not determined by ROI.)

Employee health is often the biggest factor in determining ROI, Halliday says. "ROI is much more than the direct saving of dollars," he says. "It can be much more indirect, such as not having people out on sick leave." An RFID system can often be engineered—or reengineered after a pilot—to enhance ROI by safeguarding employee health.

A recent project in a consumer industry was adapted to achieve just such a goal, Halliday says. Items shipped in returnable plastic containers must be checked for correctness upon receipt, but the containers are difficult to open, despite the use of special tools, and the repetition was taking its toll on employees' health. The new process uses RFID readers and tags to ensure the contents' accuracy so the majority of containers do not need to be opened and reclosed. The number of containers that must be checked manually has fallen from roughly 1,500 per day to less than 50 per day.

"The change started out because there were three people out on sick leave because of repetitive stress injuries," Halliday explains. "We simply eliminated that stage of their process, so they don't get that repetitive stress injury any more."

In addition to getting buy-in from top management, key stakeholders must always be given the opportunity to voice their support, or objection, before the system is fully deployed. "It's important to have buy-in from all aspects of the organization, so it's not being dumped on anyone," Spahr says. "I can't tell you the number of times I've had an angry IT person on the phone who just had a system dumped on him, and he's trying to figure out what's going on." Failing to consult project stakeholders prior to rollout is inviting trouble, he states. (For tips on getting front-line managers and other end users on your side, see Gaining Project Buy-in From Company Employees.)

Fix Things
Pilots are not only designed to show how an RFID system works, but also how it sometimes doesn't work. As deployment preparation begins, project leaders need to turn their attention toward addressing any problems or shortcomings that became apparent during the trial.

Problems can pop up almost anywhere, Lassner says. "There can be shielding issues, there can be tag issues, there can be reader issues... It's really a case-by-case basis," he says. "And, yes, you're only going to notice those things in a pilot."

A need for incremental changes during and after a pilot is to be expected on certain types of projects, says Jignesh Vania, IT head at Adani Port, in Gujarat, India. Since deploying a passive ultrahigh-frequency RFID system to manage cargo, the Hazira Container Terminal has decreased loading and unloading times (see RFID Reduces Traffic Delays at Indian Port).

"We have faced some challenges during the project," Vania says. "As we have moved cranes on which we have to mount the RFID readers and controllers, many modifications were necessary on the standard mounting accessories." Mounting accessories had to be redesigned and customized for each type of crane, but the extra work paid off. "Now, it is almost more than one year, and RFID integration is working fine without any major issues," he says.

RFID project teams handle predeployment system fixes and changes in different ways. NSWC, like most large organizations, takes a committee approach. "We use an engineering change board," Mabry says. The RFID team reviews project results and sends suggestions for upgrades and other alterations to the review panel. "Through this engineering change board, we are able to maintain configuration management and make changes," Mabry explains.

Important changes are often made just prior to deployment. The NSWC's project team, for example, thought things over and determined that its shipboard inventory-management system would benefit from the use of smaller tags. "We went with smaller tags because they are less cumbersome and more user-friendly for... systems integration," Mabry says. "And some of the newer tags have a better read range, so therefore you have a higher reliability of your inventory."

Project leaders also must remember that problems that appear trivial in a pilot have the potential to become far more detrimental in a full-scale deployment. "Companies tend to look at pilots as the least expensive way that they can test their theory, their solution, and determine whether it's going to work," Richardson says. "But what happens when there are 50 tags in proximity to a location, and you haven't considered the fact that your reader sensitivity might have to be dialed down so that you don't start picking up unanticipated reads?"

"Unfortunately, unforeseen issues will always be unforeseen," Lassner says. "It's personnel and people who make the difference. As long as the RFID professionals know what they're doing, and they've got the support of their client, and as long as the resources are strong, it makes all the difference toward minimizing unforeseen issues."

Tackle Training
As an RFID pilot heads toward full deployment, it's time to begin training the people who will be operating, managing and maintaining the system, as well as the individuals who will be incorporating the technology into their daily work routines (see How to Develop an RFID Training Program).

IT staff members and other system stakeholders need training to achieve technical ownership of the solution, Spahr says. In addition, training key supervisors and other top staff members is vital "so you can have super-users who can then train regular users," he adds. This way, companies can save time and money by solving routine operational problems in-house rather than turning to outside help.

Halliday agrees. "Once your initial people are trained, they kind of become the onsite experts," he says. "We spend a lot of time training people in the early days, because we find that's the best way to guarantee success."

Good training also helps managers avoid the threat posed by confused and resentful workers. "What ends up happening is that you get 500,000 suggestions telling you why the system doesn't work," Halliday says. "Then it really doesn't work, because that's the attitude people develop when they're forced to do something they either don't want to do or they're not used to doing."

Odette Cancer Center not only had to train its staff members but also teach a never-ending stream of new patients how to use its card-based tracking system. "The biggest issue for us was the patients," Sousa says. The first patients to use the system acted as if they were using a bank card. They didn't realize they could register their location simply by waving their device near the reader. Instead, they placed it on the kiosk for a moment, then picked it back up again. "We modified our information pamphlet and the directions to the patient displayed on the screen to scan the card. But if that's the worst [training problem], we're good with that," she jokes.

"Once everything was done and all the privacy and impact issues were signed off, and all the user groups were signed off, then we were ready to go," Sousa says. "We just decided to bite the bullet. We have to go live with this. We know what we're going to be dealing with, and there will be some hardships, but if we keep delaying this, we're never going to get off the ground."

Odette Cancer Center found its decisive moment. Since rolling out its patient-management solution, the center has achieved its goals, increasing patient throughput and shortening the wait time for patients receiving services.

Editor's Note: In our January/February digital magazine issue, we'll examine best practices for rolling out an RFID system at other locations. And in our March/April issue, we'll look at how to add applications to an RFID infrastructure.