When you grind up a title such as Pensacola Defense Switch Network Integrated Management Support System into verbal hamburger, the resulting acronym is PENDIMSS. When Pensacola's Telephone Services Department, headed by Jack Henderson, coupled government-owned, contractor-maintained DIMSS software with their concept of a Network Management Center, they created PENDIMSS. The objective of the PENDIMSS prototype is to prove that multiple telephone switches can be managed effectively and efficiently from a central location.
Management has many different meanings in a prototype such as this. In this case, although PENDIMSS routinely grinds out the familiar types of reports common in a good management tool, it will someday be capable of more futuristic applications. According to Don Howell, head of the Consolidated Base Level Technical Division that's in charge of PENDIMSS, "The company we're dealing with is working on a feature that will allow us to actually program telephone switches from a centralized site."
But, that will come later. Right now the goal is to prove to observers that PENDIMSS' administration, asset management, accounting, planning and network management capabilities are pretty spectacular.
PENDIMSS FunctionsAccounting
The PENDIMSS prototype includes NCTS Pensacola's System 85 switch, plus the switches in Meridian and Whiting Field. The prototype runs on an ethernet hub which is connected to an Internet gateway. The workstations are Sparc Classics. Every hour on the hour, dial-up modems call the switches to download data to the PENDIMSS workstations. "We're using dial-up modems because it's cheaper and easier, and we wanted to get the prototype up and running as soon as possible. As an added bonus, since we're connected to the Internet through a gateway, GTE has access to do their software upgrades," says Howell. See Figure 2
PENDIMSS isn't just another management tool. This tool does something that heretofore couldn't be done: It identifies all the costs associated with a particular telephone switch giving management the ability to devise methods of recouping some of those costs.
A telephone switch is essentially a computer designed to route and connect calls. It has all of the circuitry needed to make your phone call a reality. Obviously, switches vary in size, capability and use, but they all require care and feeding. Add a phone and the switch has to be programmed with the appropriate number; remove a phone and you have to program that change. All of this takes management, manpower and money.
One of the primary reasons for PENDIMSS is that although each company that leases or sells a telephone switch to the government has its own telephone management system, each one of them is different. Since the government has switches from many different companies, there was no way to centralize telephone switch management to determine where the waste areas were.
Some of the switches are old, some are new, some are leased and some are owned outright. The problem comes with the variety and the age. In fact, the problem of an aging switch affected the PENDIMSS prototype from the very beginning. According to Howell, "Our System 85 is an old switch. AT"T didn't have an interface because that switch used a protocol that had been developed years ago and hadn't been documented. The upshot was, we had to pay AT"T a chunk of money to give us enough information to get into the switch. We had no choice. We needed that switch interface before we could set up the prototype that included our switch and the ones at Whiting Field; Meridian, MS; and Corpus Christi, TX."
The cost of a base telephone system is tremendous. Obviously, the need to economize heads the list of reasons why NCTS Pensacola stood-up the PENDIMSS prototype. Identifying areas of potential savings is a primary goal.
One of the more visible cost areas is billing. Telephone companies send monthly service bills to government agencies just like they do to your home. Before PENDIMSS, Navy activities didn't have a way to reconcile those bills. Now NCTS Pensacola is gathering the switch data themselves and according to Howell, "Although we haven't gotten very far into this test yet, we expect to find and resolve some billing discrepancies. I don't think there will be any major discrepancies, but if we solve the little problems, we'll save a lot of money."
Other areas are equally ripe for saving money. "You'd be amazed at the number of people in a command who have a dial up modem on their desks that requires a dedicated telephone line, and they never or very infrequently use it. Those lines are expensive. If we can identify things like that, we can save a ton of money," comments Howell.
Personnel is another potential area for savings. Currently, the average site probably has four or five people taking care of the switch, moving and installing telephones and designing telephone installation for new construction. Some of the larger bases might have twice as many people to maintain a switch. It's a possibility that site support could be cut significantly by doing most of the work from a central location.
Although the potential for savings in personnel exists, Howell feels, 'We'll always have people located in close proximity to a switch to talk to the customer up close and personal. For example, our prototype manages the switch in Meridian, MS. That's more than a four-hour drive; we'll probably always have at least one person at that site to interface with the customers and move and install phones. Whiting Field is around forty miles from Pensacola. That's not too bad a drive. Some one could make it in 45 minutes. We'd need a very minimal crew there. However, I can't predict savings from personnel draw downs because I don't know how many folks there were at the APTS sites we took over. Actually, personnel savings isn't what I'm really looking for. Efficiency of billing and use will show plenty of savings by itself."
The PENDIMSS is actually made of several different pieces. The DIMSS software was written by the government and is owned by the Defense Information Systems Agency. DISA, however, has contracted with GTE to perform software maintenance and installation.
Pensacola evaluated different modules of the DIMSS software during the early stages of the prototype development. "We looked at the software and liked it. Best of all it was free. At least the software was free; installation and maintenance had a healthy price tag," comments Howell.
DIMSS software didn't come without problems. It has to communicate with a variety of telephone switches, and they all speak different languages. Part of GTE's job is to develop those switch interfaces. According to Howell, it made sense to have GTE install the interfaces they'd developed.
DIMSS gathers trunk data and is specifically tailored for each individual switch. Analyzing the trunk data can determine which trunks are getting too much use and which trunks are only idling.
The system can also identify illegal calls. It's certainly a temptation at times just to pick up the office phone and call a family member who lives out of state. Who's to know, right? Not any longer. PENDIMSS records the destination of each phone call and knows which handset was used to originate it. Howell's team isn't in the policing business. A little light doesn't start blinking each time someone places a call to a personal number and bills it to Uncle Sam. However, the information will be provided to managers who can do as they wish with it. If a supervisor becomes suspicious of the calls that show up on the monthly report, he can pick up the phone and call the same number. Management has the say on which phone calls must stop.
The variety in the switches poses other problems. The System 85 is designed to put out data once an hour. Obviously, when you're doing fault monitoring you have to remember the data is an hour old when you get it. Other switches, such as NorthernTelcoms are different. Data is available every five minutes. Why the difference - age and architecture.
Leased switches pose another problem. The port that PENDIMSS uses is the same port that the vendor uses for maintenance. Obviously, only one has access at a time. The Network Management Team has an agreement with the vendors to share access. "We don't tie it up very long - only a couple of minutes each hour," comments Howell.
Other savings are likely to come to light. It's conceivable if this system were used DoD-wide that severely underused switches would be identified. "At the very least" according to Howell, "we'd find trunks that weren't being adequately used."
Centralized management has other benefits. According to Marlow Langham, Communications Specialist in the Consolitated Base Level Technical Division, "Centrally managing telephone inventory is a prime target for saving money. Something as seemingly insignificant as the bulk purchase of telephone handsets would reduce the cost per unit to just a few dollars." Multiplied across DoD, the savings become significant.
Based on the concept developed by NCTS Pensacola, every switch in the Navy could be managed from one location. That's the theory. See Figure 3. But since no one has determined whether a region will ultimately mean two or 200 switches, no one knows if it's practical. In fact no one's sure if it's feasible to manage all the regions from a central location.
The object of the Pensacola prototype is to prove multiple telephone switches can be managed from one central region and that those switches can be taken care of with a minimal number of people physically stationed at the site.
If all the switches can't be managed from one location, there's still a lot to be gained from managing multiple switches from one location.
About the Author: Diane Hamblen
