In the past seven years, the U.S. Army has cut average domestic delivery times from 15-20 days to six, slashed order and ship times for repair parts by two-thirds worldwide, and as a result, saved billions of dollars.
How did it happen? A big part of the answer is a quiet revolution, fomented by a handful of defense analysts and Army logistics officers who bet they could take the best supply chain practices of corporate America and adapt them to the vast and complex problems of military fulfillment.
The effort began just after the Gulf War. Although the war was seen as a huge logistical success — after all, how many organizations can move the equivalent of the population of Alaska to the Persian Gulf and back? — Army logisticians say it also raised questions: How could they improve efficiency? (For instance, 28,000 of the 41,000 containers sent to the Saudi desert had to be opened to determine their contents.) What would happen if next time they faced an enemy who wouldn’t wait quietly for 100 days while they built what the Army likes to call its “iron mountain”? How could they respond now that conflicts were happening in every corner of the world, not necessarily in places convenient to the big stockpiles in the U.S. and Germany?
To add even more pressure, in 1992 the General Accounting Office placed the Defense Department in a high-risk category for “waste, fraud, abuse, and mismanagement.” Reporters wrote gleefully about a 566-year supply of hospital jackets, a million dollars’ worth of new engines left outside to rust, and billions in goods and parts that were apparently missing in action.
A modernization program began taking shape almost immediately, but for the most part it didn’t seem to matter. “We had a lot of stuff in the hopper, but it didn’t seem to be making any difference to the guys on the ground every day,” says Thomas J. Edwards, deputy to the commander of the U.S. Army’s Combined Arms Support Command, based in Fort Lee, VA.
In the Army, the top brass turned to the RAND Corporation for an answer. In 1994, analysts from the federally funded RAND Arroyo Center in Santa Monica, CA, met the Army’s 30 top logisticians in Washington and told them that the only solution to their problem was a revolution. Not with technology, although technology would have its place, but through a new way of looking at military logistics that substituted the “mass” of stockpiles for the “velocity” of rapid resupply.
The analysts dubbed the process “velocity management,” which sounded better to the military ear than just-in-time. “For a war fighter, the term just-in-time in the fog of war is pretty scary,” explains Lt. Gen. Daniel G. Brown, deputy commander in chief of the U.S. Transportation Command. “You mean to say I’m supposed to go into combat and within two minutes of going into combat my parts and munitions are supposed to arrive?”
The end of the meeting was a consultant’s dream. “Whether the conditions were right or the case was so compelling that the RAND analysts made, there was a widespread and deeply felt agreement that we ought to launch this velocity management initiative,” says Edwards.
Cast in stone
After getting the thumbs-up, RAND analysts went hunting for more lessons to draw from all kinds of logistical systems. They studied such logistics heavyweights as FedEx, UPS, and Dell, as well as leading firms in a variety of industries. One of the most useful, says RAND analyst Mark Y. D. Wang, turned out to be the toy business. Delivering Tickle Me Elmo dolls to Toledo might seem like a different business from sending anti-tank missiles to Afghanistan, but according to Wang, the toy industry and the Army share some similarities. First, they tend to have huge spikes in volume in a very short time, with the toy industry doing 75% of its business at Christmas; and second, they must respond to shifts in demand immediately, because it’s difficult to predict in advance which items will be “hot.”
But RAND analysts saw velocity management as more than simply copying industry best practices. It was a process improvement methodology, a kind of khaki kaizen. To execute the mantra of this part of the program (“define, measure, improve”), they recommended two levels of committees that would continually evaluate every aspect of Army logistics: process improvement teams, which would look at the overall structure, and site improvement teams, which would examine how things worked at individual facilities.
First, the process improvement teams got to work. Their job was to “change Army policies that were stupid [and] that impeded effective working at the local level,” Edwards says, as well as to negotiate more productive relationships with suppliers and government partners. These committees measured requisition performance times and identified some of the more bone-headed problems, such as a signature requirement for big-ticket, high-priority items that resulted in even slower delivery times than for low-priority goods. Even computerization was no guarantee of efficiency; they often found people carrying around a disk with a requisition on it. “You might as well be carrying an engraved stone,” says RAND analyst Rick Eden.
And they measured performance — or the lack thereof. When they first showed results, people wouldn’t believe them. “There was a lot of finger-pointing. The retail side would say, well, the wholesale system is very slow. Then the transporters would say, well, the warehouse guys are very slow, and then we showed delays everywhere through the system,” says RAND’s Wang.
The second part of the program, deploying site improvement teams, took more time to set up. Says Edwards, “We had a pretty good idea of how we wanted them to work, but we had a lot of difficulty, once trained, to get a senior advocate at each installation to take responsibility.”
Army bases at that time were not bastions of proactive management. Eden says that you could visit any base, just point at the first thing you saw, and see the “mass paradigm” in action, which in his description sounds like an old comedy routine about the Eastern Bloc. “You’d get out of a car and see a two-ton generator sitting outside a building not attached to anything,” Eden says. “‘Well, why is it here?’ You might ask. And it turns out nobody knows why it’s there, it’s been there quite a while. Then someone would say, ‘Somebody must have put it there; someone’s probably going to come pick it up.’”
‘Clueless about logistics’
The first velocity management volunteer was Col. James Pillsbury, of Fort Campbell, KY, commander of the division support command for the 101st Aviation Regiment. Col. Pillsbury — now Brig. Gen. Pillsbury, commander of the Defense Distribution Center — says it didn’t take him too long to decide to help. “We were, at best, getting order ship time, as we called it back then, in the 15-20 day mark,” Pillsbury recalls. “My bosses above me, who were clueless about logistics, would only ask us questions on the periphery, like ‘You got good requisitions?’ But I knew that there were inherent problems with the system.”
Over 18 months, Pillsbury cut delivery time to Fort Campbell by 300%. Setting up a regular truck route from the huge eastern supply depot at Susquehanna, PA, was a big part of the solution. Before, trucks had just arrived willy-nilly. Now, he set up a regular route, with definite stops on the base. Thanks to that truck, they were able to put the “schedule back in scheduled maintenance,” he says, save money by not stockpiling, and save the time of many soldiers and officers who would have wasted hours and hours every day just waiting for trucks to arrive.
The process also changed the ethos at the base, according to Pillsbury. One warrant officer promised him that they were now trying to unload the trucks with “a pit-crew mentality.” That awareness of one’s own place in the larger system may be one of the most lasting effects of the velocity management initiative, according to analysts. “It’s not evident to someone who’s moving material through a warehouse that it really makes a difference whether or not this box moves today or tomorrow,” says RAND’s Eden. “One of the things that measuring does when you measure end to end is you can show everyone what their contribution is, so you can show everyone how important it is that they get those details right.”
If velocity management made processes easier for the crews at Fort Campbell, it also made them easier for the Defense Logistics Agency, the base’s main supplier. Analysts say that velocity management has trained the army to be a demanding customer — and that that is a good thing. It’s much easier to respond to articulate customers, Edwards says. “It’s very difficult to negotiate with someone who doesn’t quite know what they want, has no quantifiable performance standards, and can’t really stay engaged.”
Velocity management has been good to Pillsbury and its other proponents in other ways as well: Defense officials say that virtually all the army’s logistics generals are people who took a risk with the new idea seven years ago.
One of them was Lt. Gen. Brown, who added a third star to his shoulder when he left the army’s Combined Arms Support Command to take a post at the DOD-wide U.S. Transportation Command in 1999. Brown is now pushing the idea again, this time in a new guise. He’s renamed it the Strategic Distribution Management Initiative, and he said that his idea is to use the velocity management concept and apply it to transportation throughout the Department of Defense. Brown says the effort is cutting delivery times to American units all over the world. Already, he says, his agency has repositioned 50,000 stock lines and cut ship times in some overseas routes from 40% to 60%.
But Brown finds that at this global level, while big gains are still possible, the game is much harder. “The issue is, how do you pull all that together so that you optimize the whole system even if it may mean suboptimizing a particular node or a particular budget or a particular activity?” he says. “That crosses service lines and command lines and different budgets. Trying to do that is tough.”
Bennett Voyles is a business writer based in New York City. He can be reached by e-mail at firstname.lastname@example.org.
It’s easy to think of the military as a place where improvements are accomplished with a stern order and an obedient salute, but the veterans of the velocity management campaign say that the keys to their success were more complicated than that. “Change is tough, I think it’s the toughest thing we do,” says Lt. Gen. Daniel G. Brown, deputy commander of the U.S. Transportation Command. But here are a few tips on how to go about setting up a process improvement team of your own:
Choose the right standards Brigadier General James Pillsbury says that part of the reason for the campaign’s success was a careful choice of which measurements would be monitored. Once it became clear delivery times were being cut and resources saved in the process, the team defused a lot of potential opposition.
Show the stakeholders what they have to gain By explaining to aircraft repairmen that cooperating with the initiative would mean better access to spare parts, and to the generals at his base that guaranteed delivery times would save them money and leave soldiers with more time for other duties, Pillsbury says that he earned a lot of cooperation.
Don’t meet for meeting’s sake Regular meetings are important — but don’t have them too frequently, advises RAND analyst Rick Eden. “There has to be enough time that the lower-level teams have something to report. Otherwise you’re just constantly pulling up the plant to check the roots. You have to let it grow a bit.”
Think about the customer Keeping the customer in mind is essential in making process improvements, according to Pillsbury. “As an aviator, you do backward planning,” he says. “You start with your objective and you work back to the launch. And I think velocity management did that too.”
Optimize for the whole system, not the parts “There are a lot of well-intentioned organizations and people who are trying to maximize their particular piece of the supply chain management system, but one of the things that’s come clear is that it’s not just a matter of maximizing a node,” Brown explains. “You have to be able to think end to end.”