"We don't own oil wells," says Perry Gard, VP of ERP development at RPC, "we just help the people who do to run them more efficiently." In early 2003, the "efficiency" claim was one that RPC would have had trouble making about its own affairs. An oil and gas services company with corporate headquarters in Atlanta and district offices throughout the country, RPC found itself continuing to use a paper-based order and invoicing system in an industry increasingly focused on automating such processes. To order equipment and bill clients, mobile workers recorded information and signatures on a form that was then sent back to the district office and sometimes rerouted to corporate headquarters or billing offices located in Louisiana. The costs were enormous: hundreds of man hours and extreme delays between job
completion and client billing.
Following a 2003 initiative to modernize the mobile effort and control both order entry and transactional data, RPC Project Manager Roch Johnson evaluated several possible solutions, always with two major criteria in mind: job site connectivity and worker computer literacy. The possibility of remote connectivity at most job sites was eliminated early: "A lot of these sites are very remote," says Johnson. "Oddly enough, we got some of the best reception at the offshore sites [in the Gulf of Mexico]." Understanding that a workforce consistently connected onsite was an unrealistic goal, RPC cycled through several options before employing the services of Atlanta-based Optimus Solutions. In late 2003, Johnson presented a detailed specification write-up, complete with mock screenshots, to Dale Ewald, a consultant with Optimus.
Johnson's meticulous preparation was rooted in RPC's need for the solution to be extremely user-friendly. RPC field workers fall into one of two distinct categories: well workers responsible for site maintenance and repair, and engineers charged with more complicated work. The latter group was, on the whole, experienced and receptive to technology; RPC was not as lucky with the well workers. "I remember during one of our initial usability tests," says Johnson, "we were trying to explain the technology to one of the well workers. He told me he had been on the Internet once in his life."
Johnson's program write-up was based on existing online technology already in place at RPC. The challenge for Ewald was converting the online browser form to a version that could both operate independently of an Internet connection and could be instinctive for a large-scale group of workers untrained in the technology. "We basically had to first build an online program, then find a way to convert its capabilities offline," reports Ewald. The solution to this problem came in the form of Domino Offline Services, an IBM/Lotus utility that allows Web applications to be taken offline.
Another significant shift in software came with the move to Lotus, and the use of the Lotus Enterprise Integrator. Johnson cites Lotus' proficiency with replicated technology and its simplicity in controlling the user experience as huge positives in the solution. While the layout of the newly dubbed "Mobile Equipment Service Order" software was strikingly similar to existing software used by RPC, the changes were instrumental in the shift from browser to offline application. One of the program's features of which Ewald is most proud is akin to Google's Ajax technology, which allows users to look up server data without sending an entire form. While admitting that Google's version is faster and more complex, Ewald also says, "Up to that point, the RPC application was the best solution I had ever developed. We put together a heavy application in a very, very short time frame."
There was a limited deployment of the technology as usability testing began in the spring and summer of 2004. RPC equipped a small number of workers with Dell D600 laptops and Topaz SigLite electronic signature pads with a USB interface and a simple Windows XP operating system. Gard says initial reactions from the workers were mixed: "It was difficult for some of the workers to see the full benefit of this technology because they did not see the efficiency improvements on their end of the job." Large-scale training was also a problem; with workers spread throughout the country, many of them at remote sites, the technology could not be distributed with training in one fell swoop. A trickle-down model, where administrators were trained and expected to pass the information to their workers, was attempted first with limited success. Both RPC and Optimus admit that training the mobile workforce has been a difficult task.
Training continued, though, and as the program was fully deployed in December 2004, the problem of transferring recorded information from the job site to the district office re-emerged. Although some of the workers in less remote areas were provided with wireless PC Cards, RPC did not retain a national vendor; most synchronizations occured a few times a day when a worker had a chance to connect to the Internet via landline. Commonly, a worker only connected once per day, from a hotel room after he had completed his daily work. Because of travel, workers often found themselves outside of familiar areas and had difficulty connecting directly to the office. To counter this problem, Johnson and Ewald equipped each laptop with iPass, a simple program that manages the phone line setup to the district office once the computer is connected. The final piece in the puzzle, iPass enabled workers wholly unfamiliar with Internet connectivity to deliver their info to the office with relative ease.
While discussing the deployment in retrospect, RPC and Optimus acknowledge that, given the opportunity, certain aspects could have been handled differently. Johnson admits that the worker training could have been initiated more uniformly, and that certain aspects of the Lotus software initially made data extraction difficult. Results from the technology have, however, been overwhelmingly positive. The client billing process, which initially took 45 days, now takes an average of seven. Return on investment occurred in less than six months. And the application, now in version 1.3, with 1.4 on the way, has continually been streamlined for user familiarity. The company whose business is efficiency can now boast of some of its own.
Adam Brickman is a writer based in New York City.