Railroaders' Nostalgia > The BN[SF] Kansas City Hump Debacle

Just as background, because the background is a long story itself, I was one of 36 BN train dispatchers assigned in 1991 to develop operating solutions to system wide congestion and service failures. We were hand-picked for experience rather than assigned by seniority. The two criteria are similar, but there were some lower BN seniority boomers in the group, like me. These assignments came from the top, which upset our immediate bosses no end.

For me, that led to capacity analysis and planning capital projects, service design, and to developing passenger service projects, notably the Seattle commuter service and the Washington State rail program, but also involved work on potential commuter service in Houston, Dallas-Ft Worth, Vancouver BC, and Kansas City and potential high speed passenger service between Dallas and Houston.

Management kept telling me that the assignment was temporary because I was not promotable (degree required but mine was the same degree as Gates, Jobs, et al.). I had to have a degree to do what I was doing officially, so I had to be doing it unofficially as a temporary assignment. I remained a train dispatcher (union position) throughout (which came with its own set of problems). Along about 1995, it was obvious that all dispatching would be moved to Ft Worth. I repeatedly told the bosses that I would not go there, I was staying in Seattle. I was told that I would be pulled from my temporary assignment anyway and be put back to "pushing buttons." I wasn't about to go back to what is now called train dispatching no matter where it was and was very direct in relating that to them.

Well, the day came that everybody left except me. They didn't believe me. They overlooked giving me a notice of abolishment of my position and assumed that I would just go with the rest of the office. I didn't say a thing. Here I was by myself in Seattle on my temporary assignment. I was told that was redundant under the new regime, so I was assigned to the TSS (Santa Fe Information System) cutover. I was to be part of a team of train dispatchers, yardmasters, and some others, from both railroads, to go around to every BN terminal at cutover time to teach the yardmasters how to use the new system.

It wasn't pretty. We had a short training session in Kansas City. We only needed to know a few basic commands because that was all that anyone needed to know (like the BN COMPASS system, for which I was told the same thing when I hired out, there was a big manual of stuff nobody needed to know). All we had to do was spend a couple of hours teaching this handful of commands after cutover and move along. The plan was two people in town for seven days, two days off at home, and on to the next one.

It didn't work out that way. We wound up double and triple teaming and staying in one terminal for weeks. I started at Willmar MN. It didn't go well. I went on to Grand Forks ND. There was a Santa Fe guy and me. The place cratered on cutover day. We worked 16-18 hours a day for a week trying to troubleshoot at the same time as we were teaching an expanded range of system operation to yardmasters and managers. The few basic commands we were supposed to teach them couldn't handle what they were doing in the yard. They needed others to represent the operation. But (Surprise!) a lot of the ones we were told not to use because nobody would need them anyway didn't work. Some did nothing; some made the original problem worse. We did a lot of mudhopping (walking tracks to list them up on paper). They brought in a second team. After about three weeks in town, it was under control (remember, when the fire department says a fire is under control, it isn't out). My Santa Fe colleague and I were sent home for a couple of days, then on to the next adventure.

Birmingham, Springfield, Great Falls, Spokane, Missoula, and Seattle were similar experiences. We heard through the grapevine of the team that Memphis, Cicero, and Northtown were a complete disaster and the cutover would be stopped for a few weeks while we caught up.

Kansas City, BN Murray Yard, was the most memorable of the entire assignment.

I was sent there to relieve the original team. They had been there for three weeks and the system still wasn't working. There were some big problems.

I showed up in the hump tower for my 12 hour 6p-6a shift. As soon as I took the transfer, the guy I was relieving was going home. There was a pile of lists on the desk. They were losing cars; a lot of them. The hump lead had over a hundred extra cars. Nobody could figure out how they got there. I told the guy to get out while the gettin' was good, before they did like happened to my team a few times already: cancel your return trip, you can't leave.

Night after night, the hump lead list grew and grew. I was glued to the lists, the window facing the bowl, and the computer monitor. Why are these cars disappearing from the bowl and winding up on the hump lead? I found it.

The hump computer and the TSS system were separate and had only one way conversations. The hump computer could update TSS, but TSS could not update the hump computer (you IT guys can stop rolling your eyes now, it gets better). The logic of the hump computer assumed that all actions were sequential. The first car goes over the hill, the first car clears into the track, the second car goes over the hill, the second car goes into the track, and so on until the whole cut is humped.

The real world doesn't work like that. It is possible, and not unusual, for one car to pass another, clearing in its track before the car ahead of it does. Think of a 40 track bowl with the hump lead in the middle, straight into 19 and 20. The distance between the hump and 19-20 is a lot less than the distance between the hump and 1 or 40. The programmers didn't know that because they knew computers, not railroads.

I watched in fascination for a while as I would see a 20 shot clear before a 5 shot. The 20 shot moved into 20. The 5 shot stayed on the hump lead. I watched for a while to be sure of what I was seeing. There was no doubt about the hump computer logic. I also figured out the one way communication. I watched as trains arrived. The AEI arriving moved the cars responsibility from the train to the yard in TSS and in the hump computer. As the status and location in the yard changed in the hump computer, it did in TSS. The outbound AEI merely created a new record in TSS for the cars missing from the train outbound list from the hump but didn't remove them from the yard.

There was a way of creating a temporary record of a car in the hump computer. The clerks were doing that when they listed up outbound trains. Those temporary records didn't move the car record into the train, they only put a placeholder with a car number in the train. As long as there was no action on the real record, the placeholder would remain. If the yardmaster moved a car that had a placeholder somewhere else, the programmer's assumption was that they yardmaster knows all. They didn't know the difference between marking up a list and listing up a track.

The special help desk assigned to the team had grown tired of me reporting problems during the meltdowns I had the misfortune to experience. They had become so tired of me reporting problems in detail (which they didn't want to hear) so that they could troubleshoot (which they said was not necessary because it was a user problem - me - and not the system) that they wouldn't answer the phone when I was working.

I was there for two weeks. It took the first to figure out what was happening. For the second, I came in every night and watched the impending debacle. On the last night of my assignment, it happened. Thanksgiving was coming up. They sent management folks from the IT department to relieve us scheduled folks so they wouldn't need to pay overtime in addition to holiday pay (another long story). The first management guy showed up at about 5am. He wasn't my relief. The yardmaster complained about the difficulty of working around the 300 lost cars that had accumulated on the hump lead. The management guy said that they should have had a lost track. He showed the yardmaster how to create one. I intentionally didn't do that because I knew what would happen. I didn't say anything because I was really tired of being told by the IT folks that I didn't know what I was talking about...if they would even answer the phone.

Once the lost track was created in the system, the management guy instructed to yardmaster to move the 300 extra cars from the hump lead into the lost track. He did. I wasn't asked, nobody wanted to hear, I said nothing. The computer records for the 300 cars were instantly sucked out of every train or yard that they were in and moved them in TSS to Murray Yard as well as moving them from the hump lead into the lost track. Now, advance lists were wrong, interchange lists were wrong, tracks in other yards were wrong. Even the lists of the tracks in the bowl were wrong because the placeholders disappeared.

There was third degree Armageddon in Kansas City. The Terminal Superintendent came unglued. The Superintendent came unglued. The control center managers came unglued. The terminal superintendent was on his way to the tower to deal personally with the TSS guy. That was me...for another five minutes. I had a precious day before Thanksgiving ticket home. There was an ice storm due that afternoon. If I missed this plane, I would be in Kansas City for a long time more. That was not on my list of things to do.

The guy who was supposed to relieve me, not the one who created the lost car track (he did that and left to do other greater and more important things) showed up. I gave him the Cliffs Notes version of the debacle, told him that it was all his and by the way, the Terminal Superintendent is on the way to see you...oh, there he is now, and left for the airport.

I had way more stupid stuff than I wanted to deal with. I checked in the rental car at the airport. I was asked "did you fill the car with gas?" I couldn't resist: "No.........just the tank, kept on walking to the terminal, and got out of town just before the freezing rain became serious and shut down the airport."

TAW

And that's the way things were. At the NOC in Fort Worth there was NO formal training for TSS provided. Thank goodness we at least had the former Santa Fe train dispatchers that taught us poor BN relations how the system worked. Things must not have changed much after the coming of TSS as the training new student train dispatchers got was rudimentary at best. It was interesting though in that every student I had taught me new procedures and I did the same for them. So much for formal training.

After the merger had been announced BN was still trying to re-invent the wheel on their various data systems as COMPASS was on its' way out or so we were told. I went to a 4 hour class one day that turned out to be a typical waste of time meeting. After the class convened, we were told there would not be any hands on training due to main frame failing earlier in the day. So we sat around and watched countless drawings projected on the wall of what the new system screens would look like. About halfway through we were further told that COMPASS would stay with us until the merger was approved and at that time TSS would show up.

I'm glad the new programs were never implemented as some of them looked like the amount of manual entering and updating was going to increase where at least the current processes had automatic updating! Oh well I had fun filling out the survey handed out at the end of the class.

Looking back at your problems,I am glad that I did not need anything more than
common sense to do my job.Altho,I had a degree.

Uncle Joe,retired car foreman

Do you guys ever stop and think 'how did we win the war"?
All those guys with degrees and none of them can figure out a light switch!

Good story TAW!

KR

When I was a field technician for 3M, Computer Aided Dispatch was introduced in the mid-1980s.
It was more like Computer Aided Debacle.
We (and our customers) suffered for six months before 'management' pulled the plug on that gem.
...Lorenzo

It wasn't much better trying to work the Ark City pool out of KC either. Horrible line-ups (Lie-ups) , everything set back for power.  As soon as something was ready, it went to a Wellington train. If they furnished it at all, it was the worst of what was left. The yard was so fouled up you couldn't get your power off the pit, and out to the train anyway. Coming in wasn't much better, sitting for hours at Morris waiting for a track. Even the WEllington pool guys got to enjoy that. Almost every train had a few KC cars that had to be set out, and the "fast tracks" were constantly clogged up. It made no sense to anybody to accumulate the KC set-outs out west someplace, and run them every day as a solid KC train, No, everybody had to wait to get in, delaying every train.

And to think, the UP-SP merger disaster was making the BNSF look good by comparison. The whole 1990's merger debacle really set back the entire industry just when the railroads were finally getting thier act together following the turmoil of the 1980's downsizing and the early 1990's recession.

Ted Curphey
Ontario, OR

I really enjoyed ddg's comment about "...accumulating KC setouts outside of town someplace, and running them all in on one train, instead of every train having a few cars to set out."

In 1998 I was working the Roseville to either Sparks or Portola pool, and was unfortunate enough to catch one of the RVROM's on duty about 8pm supposedly every night.  This train picked up and set out at Oroville, then did all of the one or two car maintenance of way work at various spots up the canyon.  Next came a setout and pickup at Quincy Jct.  followed by a setout at Portola.  The train usually didn't make Portola with one crew anyhow, but then they did this to us.

Due to the power shortage, an RVROM hadn't run for three days, so they had mindlessly built another one every day, with four blocks of cars to set out on each.  UP assigns cars to be picked up to a certain train, and doesn't show cars not consisted to your train on your train list, even though they are in the track ahead of your pickup!

They found six big units all at once, and ran three RVROs an hour apart with two units each.  After meeting a couple of higher priority trains enroute, the second RVRO is now a siding behind me.  When I got to Oroville, I was able to set out my Oroville cars in a clear track with no problem, because they hadn't had any inbound cars for three days.

Then I started looking for my outbound cars, which were listed in track two, which was about full with 60 or 70 cars in it.  This is an Eastbound move, and the Oroville Local does most of its work from the East end of the yard, so the oldest cars in a track are on the West end.  After walking about 40 cars that aren't on my list, I finally find my pickup to be the last 20 cars on the West end of the track.  The whole track goes East, but it is scheduled to three differrent trains in reverse order.

While I am setting out and hunting my pickup, the Dispr lines the second RVRO in behind me, and they watch me out hunting my cars.  Meantime the third train is being held at Craig, which is the first siding West.  Now I pull the whole track out and set the rear 20 cars onto my train in the siding, and put the front 40 or so back.  I make an air test, and pull up to Mitchell Ave to wait for a Westbound, so I am still blocking the 2nd train from being able to set out. The Westbound goes by, I leave, the second train pulls up to do his work, and the Dispr lines the third train in behind the second to watch him work!

I go about 25 miles and set out a carload of tie plates, then go to the next auxiliary track which has four cars in it, although only the West one shows on my list, and it is the one I have to pick up.  The other three are empty also, but they don't show to move. So, I set one out and the rest back, and head for the next work.  Shortly thereafter, I hear the second train picking up the rear 20 of the remaining 40 at Oroville,  We get to talking, and I find out he has one car to set out at Tobin where I have just picked up one, and he has to pick up what turns out to be the rear car there also. One more car to set out in a track that shows clear, but actually has three cars in it, and we head for Quincy Junction short on time.

Quincy Jct. shows 5 cars for me to pick up, and I have seven to set out, only problem is that both tracks are full of loaded cars scheduled part for each of the three of us, and we all have emptys to set out.  Since I have more to set out than the interchange will hold, I switch out my five loads, set out five emptys, and wait for the releif crew.  When the Armadillo shows up, there are two new crews, one for us, and one for the train behind us.

While we are tying up at Portola, another Armadillo driver is on duty for the third train, who sat behind us and listened to us work.

They should have switched the three trains in Roseville, and made the first one all through cars with only a pickup at Quincy Jct.  That way he would have made it, and stayed out of everybody else's way.  The second train should have had all three days worth of Oroville cars,  picked up all of the East cars at Oroville, and all four cars at Tobin.  The third train should have done the rest of the M of W work in the canyon,  set out the emptys at Quincy, and gone into Portola with ten or twelve cars.
The other thing to consider, is that all Eastbounds were limited to 5700 feet because of non-clearing Westbounds
G. 

ddg Wrote:
-------------------------------------------------------
> It wasn't much better trying to work the Ark City
> pool out of KC either. Horrible line-ups (Lie-ups)
> , everything set back for power.  As soon as
> something was ready, it went to a Wellington
> train. If they furnished it at all, it was the
> worst of what was left. The yard was so fouled up
> you couldn't get your power off the pit, and out
> to the train anyway. Coming in wasn't much better,
> sitting for hours at Morris waiting for a track.
> Even the WEllington pool guys got to enjoy that.
> Almost every train had a few KC cars that had to
> be set out, and the "fast tracks" were constantly
> clogged up. It made no sense to anybody to
> accumulate the KC set-outs out west someplace, and
> run them every day as a solid KC train, No,
> everybody had to wait to get in, delaying every
> train.

Adding to that, many of the inbound trains had relief crews sent to take the trains in, only to sit for 12 hours and taxied. This would happen several times for some trains.

Fred in Wichita