So, how to implement this in

​How do you know that Dyar's rules are correct or reflect what people want?

How am I going to use my smart phone to read my lists when I am using it as a throttle.?  Fatal programmer error, assuming your application will always be filling the desktop.

This assumes that these are desirable.  Having worked for real railroads for 37 years, they don't ship stuff randomly.  Industries don't receive cars randomly.  You said you want a system based on a shipper.  That eliminates randomness.  I don't want to have my users guessing about the traffic levels, I want the yardmasters to know and as a layout owner I don't want wild variations in the traffic.  If the local has 2 cars one day and 20 cars the next it will be a disaster for the op session.  Wild variations in work cause the schedules and connections of the trains to go right out the window.  I have designed a system where the train have about the same number of cars, plus or minus a couple cars

Inbounds

​Several things you have to look at.

​If you at just loaded moves that requires the user to find off line (virtual) sources for receiving industries.

​If you look at just the demand for shippers, it requires the user to define shippers offline (virtual).

If you look at just the demand for receivers, it requires the user to define receivers offline (virtual)

​If you look at shippers and receivers separately then you have to balance or link the two, they can't be independent from each other.  If they are independent, then if the box company ships 3 cars but the candy company only demands 1 car you will end up with 2 cars in limbo, shipped but not consumed.

Routes?

Another consideration. 

Your shipments have to include a route component.  For example, on my layout, modeling a branch line from a junction with the rest of my railroad at A to a terminus at Z, there are 7 possible route pairs for an on line industry:

1. A to industry
2. PRR interchange A to industry
3. PRR interchange D to industry
4. PRR interchange M to industry
5. PRR interchange Z to industry
6. B&O interchange Z to industry
7. Other on line industry to industry

No matter where a car comes from or goes to, it has to fit into one of those 7 buckets (14 with the other direction).

If my box company wanted to ship to a candy company in LA , it could use routes 1, 3, 5 or 6 (routes 2 and 7 connect to PRR branches that don't have a thru connection, route 7 is entirely on line and moot in this case).  Obviously it becomes important which of those 4 routes is chosen because it completely changes how I handle the car.

"​How do you know that Dyar's

That's a good point. The state of model railroad design and the hobby in general has greatly advanced since 1952. We have a lot more resources and access to information they didn't have in those days.......DaveB

@Dave - Wow!

I think "what we have here is a failure to communicate"

We might be talking on different levels.

First, let me say that to me, operating a model railroad is like solving a puzzle - how to get the cars where they are supposed to go in the most efficient manner possible. The computerized system that I've been testing helps me do that by removing most of the bureaucracy involved in card systems. It doesn't even attempt to simulate the entire freight car management system of any particular railroad, but it does try to operate within the framework of that system.

I'm not saying the Gerald Dyer's five rules are for everyone. There are many people who do subscribe to them, and there are others who disagree. There is a great analysis of car card systems in an article by Doug Smith in the December 1961 edition of Model Railroader and a search of the MRH web site brings up a lot of discussions on these points as well. If randomness and unpredictability are not for you, that's OK.

I will say that in my short time as a brakeman and then traveling agent, the entire system seemed random to me. I'm sure it wasn't to those folks involved in supply chain management (a field that I got into later) but the jobs that I worked seldom set out or picked up the exact same number of cars at the exact same interval at any industry. As a traveling freight agent, I can't think of one customer that always shipped the same number of cars on the exact same day every week. The CP local freight that passes a block from my house sometimes has two cars, sometime several dozen.

When I'm talking about randomness, I mean unpredictability within certain constraints. That's why I set minimum and maximum values on when the shipments go out and how many cars comprise that shipment. A little experimenting with these values prevents wild variations in traffic. I know from working in the logistics field that it's a lot easier to manage things if we can reduce these variations. It allows us to predict and anticipate events. That's not always the case and it's why, when (surprise - surprise) as a member of a yard crew one day we had to make three runs out to the local Kraft cheese factory because they were loading 12 cars of cheese that day and their docks only had room for six. (Trip 1 - spot six, Trip 2 - pull six & spot six, Trip 3 - pull final six) The conductor was especially grumpy that day because he had things that he wanted to do after work.

About shippers and receivers, I believe that it's a fact of life that the North American freight car fleet is always unbalanced. There are always empty cars sitting around somewhere and there is always an unfilled demand for them somewhere else. If Fred's Box Company ships two cars but only receives one of paper, it means that there will be two empty box cars someplace, maybe off-stage, and one empty on Fred's spur. I don't know that there is a problem with that. The empty off-stage cars are then either allocated to another shipment coming on to the layout from off-stage, or one or both of them sit in a hold location until they are needed. The empty car on the siding can be loaded at a nearby shipper, or it can move empty to some holding location.

Concerning routing of cars, again I think that is something that takes place at a higher level and isn't always visible to the guys throwing switches on the ground. The switch lists and wheel reports that I worked with and have seen since don't really address the complete route for a car, not that I can remember anyway. The waybills do, but we worked off documents that just told us what's in our train, what to pick up from where, and what to set out where. This is why I think we aren't communicating. I'm looking at what the switchman or yard clerk needs, and I think that you might be looking at a level or two higher than that.

Oh, and about using a smart phone as a throttle, I do that as well, but when I upgraded to my current phone, I kept the previous one. It works great to view switch lists, yard reports, etc... and it also has applications on it that read QR codes and RFID tags. The engineer's phone runs the power, the conductor/brakeman's phone handles the electronic paperwork.

Ooops, I forgot to hire a conductor

If you can imagine a throttle in one hand and the iPod touch in the other, guess where the uncoupling stick was.

Great thread, not sure of the purpose of the QR codes (RFID, etc) or how it will make it easier. (Although technology is always fun).

In my experience on my test system, I found that once the "system" placed a car in a train and that the train moved to a new location and the system was updated to that train's location,  the location of the car was known.

This repeated until the freight car was at the yard where it would be assigned to a local switcher for placement at the receiving industry. Once placed at that industry the "conductor" would update the system.

This repeated until all freight cars were "introduced" to the system. From that point on the system knew the location of all freight cars on the model railroad and controlled their movements based on updates from the crew. Unless the crew made an error. Then the system had to be informed through a correction in location of equipment.

So my question is, are the QR codes meant to assist in finding the a freight car in the yards or assisting in obtaining the information for the cars such as reporting mark, number, etc? 

Eugene

@DaveB - Dyer's rules

Good points about all of the changes that have happened in the hobby since th 50's! I wonder what some of those guys who were "operating" in the 50's and 60's would have thought about DCC, Z scale, and so on...

About the five rules that Gerald Dyer described, I think that most people who interested in the operational aspects of model railroads would agree with many of them. I can't see that someone would purposely want to implement a car forwarding system that would require lots of time-consuming paper work, or one that was not easy to use. The unpredictability, that we can talk about.

@Eugene - The Conductor is AWOL!

Good point about trying to operate solo. When it comes to switch lists, I do like to print them, probably because I'm kind of old-school and do like pieces of paper in my hands when I'm working a job. I can make pencil notes about what I'm doing. I also like to use the electronic gadgets, though, to sometimes look up things about the cars, shippers, etc... With hand-held gizmos it's easy to do that at track-side.

Your comments about cars moving in trains, their locations being updated, and so on are where I'm going with this series of posts. One of the time-consuming tasks in using other computer-based car forwarding systems is the updating of car locations. By marking both cars and locations on the system with QR codes (or RFID tags in larger scales) we can use bar code scanners or RFID scanners to do the data entry for us. This cuts the time needed, improves the accuracy (no typos) and can be done at trackside, as opposed to taking the paperwork to a PC somewhere else in the room or clubhouse.

I think that I'll skip ahead a bit in the posts that I was planning on send out and get to how I've been using this technology to make my life easier. I can get back to how the software works later.

When it comes to placing QR codes, bar codes, or any other sort of marker on the rolling stock, some folks just won't have anything to do with it because it does detract from the appearance of the models, many of which have taken hours if not days to detail and weather. I wouldn't want to pull paint and weather off the tops of my cars either. There are solutions. John Allen, of Gorre & Daphetid fame, used plastic and hard paper tags on top of his cars to help with operations, but he was also a terrific photographer and since his markers were not permanently fixed to the cars, he took them off when taking photos or running trains for visitors and didn't damage his wonderful models.

So, to answer your question about how I use QR codes, finding a car in a yard or finding out where a car is going and what it's hauling, the answer is both. The identification part is especially important to me because I'm currently working in N scale and it's tough to read those tiny reporting marks on cars in yards.

I can...

I can. I love Time Table & Train Order operation, which requires a lot of paperwork if you are being prototypical. Easy to use is subjective, but if you are following prototypical practices, every form has a purpose and they didn't design or use anything they didn't need.

My layout is a single city, entirely within yard limits, and as exact a copy of the prototype as I can manage in my basement. On the main portion of the layout I think there are two missing tracks. All of the work is dropped off at one of the two yards, and handled by two switching crews assigned to the city.

The work is managed by the Agent/Yard Master who receives the incoming paperwork from the road crews, and assigns it to the switch crews using all of the prototypical paperwork that I've been able to determine. My friend Chris had a great observation when he was the Agent. He said it was one of the best positions on the layout. Sure, he wasn't operating the trains. But unlike everybody else, as far as he was concerned, he was operating a railroad. Everything he touched or dealt with during the session was the same on the real railroad as it is on the the model railroad. When he received a stack of waybills, and dealt with the paperwork, then gave the needed paperwork back to the crew, as far as he was concerned, that crew was going to climb aboard their 44-tonner and get to work. 

When we get around to it, the Agent/Yard Master/Towerman will have a similar approach. He has a closet that we can eventually make the inside of the tower. By mounting a camera in the model tower, and monitors behind the window frames, they will be able to see exactly the Shoreline activity directly. All of the interactions with the crew will be as if they are operating the real thing.

There are a lot of things to model on a model railroad. The scenery, the trains, the train movements, the train mix, etc. And the paperwork and operations is one of the things you can model. I'm not saying that you should want to model it, but it would be great if your computer program supports multiple approaches to modeling. To the folks that don't like managing the paperwork, my recommendation is, "be an engineer." An engineer only has to look at (or sign) a couple of forms on the railroad. Everything else is handled by the conductors.

One of the problems I have with a great many of the operating approaches and aids, whether it's paper, cards, or computer, is that after all these years they are often modeled off of other people's operating systems. They are efforts to "improve" an existing system. I would suggest developing a system that simulates every aspect of the prototype in regards to car movements. You don't need to worry about collecting bills of lading or sending out freight bills. And since waybills themselves were typed, unless you have somebody proficient (and the right kind of paperwork and typewriter) I think preprinting the waybills makes sense. Beyond that, I'll work as many of the railroad jobs into my model railroad operations as I can. Because it's really interesting, and it's fun.

The movements themselves are dependent on what's happening on the ground. NY-2 is late? It doesn't make any difference. Those cars will get here when they get here. A hot car comes in? That's fine, we'll do that first. Switch crew #2 is green? Switch crew #1 will end up handling more of the work naturally. 

I have a general idea of how things will operate, and we give very detailed instructions about how to operate the railroad. On the prototype, a crew has to be qualified on a line or job before they can run it. So they will learn the tricks that simplify a particular job from guys that have already done it. But despite that, everybody does things a little differently.

It's a complex, but interesting puzzle. I prefer that the computer program doesn't assign specific cars to specific trains, other than the initial trains, because things can change. Crews will build and block the trains according to the car routing rules as directed by the agent. The order of cars in a given block doesn't matter, but the blocks do. If cars aren't pulled in time for the first train heading in that direction, they'll make the next one.

What it really comes down to is that part of what we're modeling is the paperwork of the railroad itself. If your system can't support that, then it's not going to work for us. This is the main reason I haven't found a computerized system that works for me. They are too limited in their scope. One that encompasses the widest picture possible, but allows the modeler to choose what features to use or not is what I'd recommend.

Randy

QR Codes

I'm glad you posted this. I was wondering where the QR codes came in, although I suspected it was something like what you're describing.

I like this concept from the standpoint that it means that the actual car location is known in real time, and doesn't rely on predetermined car movements to be followed correctly. It would also allow you to design a more prototypical car movement into the system, since the movement is now car dependent, and not train dependent like other systems.

One of the things I've been trying to figure out on my layout is how to handle the movement of the cars within the yard (and yard limits). In my era, both car cards (that were stapled to the side of the car) and chalk marks would have been in use. I have a freight agent managing the paperwork, car orders, and movements, who is also acting as a Yard Master/Yard Foreman for movements within the yard. I believe, though, that a lot of those movements would have been handled by the crew based on the chalk marks and/or car cards.

I'm not sure I like the idea or practicality of putting these types of paperwork (or codes) directly on the car. Sure, if you're modeling an era that used bar codes, you could probably make an operational system - at least in HO or larger). But any time you move those notations off of the car themselves, you add complexity (and paperwork that didn't exist prototypically). So I'm very interested to see how it's working on your layout.

Another factor, of course, is the feel. So I'm looking for something that feels appropriate for the era. Also, I guess I'm going to have to see if the computer program is sophisticated enough to cover what I'm looking for (most of which I've noted in other comments). The only purpose for the QR codes is to keep the computer up to date as to where the car is located. So if I'm not using the program, I won't need those.

Randy

 "The only purpose for the QR

No matter what program one had being able to track the location of cars in real time would be handy, especially for fun running between operating sessions. It would be very handy to have the computer know where everything ended up and adjust it's data accordingly.....DaveB 

Granularity?

What granularity are you expecting?  Station?  Track?  Spot?

The answer determines the number of readers you would need.

@Randy - Paperwork and puzzles!

Thanks for the post. I agree with all of it. When I first started out in HO what seems like a million years ago, I typed up and made copies of my own blank switch lists, train orders, and so on. When I talk about paperwork being a burden and taking away from the enjoyment of operating, I don't consider what you are doing as bureaucracy. The stuff that I consider a burden are things like checking car cards to make sure the cars are in the right place, reorganizing the waybill cards so that they are in the right place, things like that. That type of paperwork has more to do with the mechanics of the car forwarding system in use than the actual documents needed to run the railroad.

One of the clubs where I was part of the crew had several different groups of guys, some interested in operation, some in detailing, some in track construction, and so on. Every so often we'd meet for an operating session and find cars rearranged due to track maintenance, or new structure installation. Those tasks needed to be done, but an unanticipated side affect was that the whole evening was spent in "putting the railroad back together." 

You are right - it is an interesting puzzle!

"operating a model railroad

"operating a model railroad is like solving a puzzle - how to get the cars where they are supposed to go in the most efficient manner possible. The computerized system that I've been testing helps me do that " 

So if the computer is working the puzzle are you just a slave to the computer? I lke the idea of solving the puzzle my self.

I have one layout I operate on were you get a general list of instructions on what your job is but it up to you to decide how it is done.

Another railroad has an entire check list you most follow to the T.  This box car must move at this time then this car then, this then, this and only in this order at this time.  You have no decision making of what to do all you are doing is moving cars from spot to spot as per a list and do not mess up the order or the entire system will get sucked into a black hole and you do not want that.

I enjoy the 1st layout and operate on it frequently as I learn things about the jobs from past operations I can improve and figure out the so called puzzle.  The later railroad nothing to learn other than follow the same list over and over every time nothing changes always the same Year, month and day are played out every time.  No need to figure out nothing as you have instructions on what order, what cuts what to move to what location in your train the car pull and drop of most be done in this order so at the end of your run your train can be check to make sure all cars have been picked up in the exact order and place in the train in the exact spots.

As an operator what layout would you rather work on?  One that use your brain or one that every thing is done for you?

"As an operator what layout

You don't have to choose either method, there's a sensible middle ground where train crews decide on the moved required to do the job, just like they do on prototype railroads.Of course the era and location modeled has a big effect on the amount of crew discretion but most folks figure that out while deciding what era and place to model  ...DaveB 

Prototypical

Actually those things are "prototypical" (I had to do the equivalent on a real railroad and have documents from real railroads in the past where they did that) since real railroads don't have a QR system like you are proposing.

​

​

Car Assignments

Assigning Cars to Industries shouldn't be that complicated with a little thought beforehand. Out on the prototype we refer to car movements in O/D pairs or Origin/Destination pairs. The Marketing and Sales Department go after the business and establish the rates for the movement. Whenever more than one railroad is involved in the O/D movement a mutual agreement has to be made where to establish Interchange if there is more than one connection between the two carriers.

In the event that a customer is located fairly close to an active Interchange with another carrier often that customer may be set up as one that is Open to Switch. In this event the home railroad may only get a switch move while the other carrier may be able to obtain the line haul. Railroads try to limit this open access while customers love it because it allows for a lower rate, especially if both carriers serve the same market. Another form of movement is called a joint rate. In these instances the two carriers that will handle the shipment typically agree on an interchange point around the halfway point between the O/D pairs. This is why UP or BNSF may choose Chicago or St Louis as an Interchange with CSX or NS instead of New Orleans for traffic destined for the Northeast.

The big factor a model railroader should figure out is which car type and car owner would most likely show up at the industry and then devolp a pool of cars that could naturally show up at any time. Normally customers try to stay pretty regular requesting x amount of cars daily, every other day or twice a week. Once a RR marked car is spotted for loading or unloading they have 24 to 48 hours to release it without being charged. Private cars can stay forever.

Another factor is to decide whether the customer is an open gate or closed gate customer. For open gate customers cars are spotted on arrival and only won't be spotted when the track is already full. That is when the cars will go on demurrage. For closed gate customers cars show up at the serving yard and remain there until ordered in. At that point they are Available for Placement or put in AP status. 

Tim

@Tim - Filling car orders

Thanks for the explanation. I had found parts of what you just posted on a couple web sites, one belonging to the BNSF and the other to CSX, but your post cleared up what was missing in both. Probably just by luck, the system that I'm testing does involve the idea of origin/destination pairs and the concept of some number of cars being loaded at certain intervals, such as daily, weekly, etc...

I'm guessing that the open gate is tied in with what I remember as "constructive placement", where a car arrives but is unable to be spotted due to something on the receiver's part. I ran into this when working as a traveling freight agent in the 70's in northwestern Wisconsin. Some of the pulpwood shippers wouldn't have the cars on their spur loaded by the time that their next batch of empty gons showed up. The new cars would be marked as "constructive placement" and sit somewhere nearby until the loads were pulled. If I remember right, the shipper had one or two days before demurrage would kick in.

This happens when operating my test railroad. Four of the six locations to spot cars have room for only one car at a time. (This is a small 1' X 4' switching operation.) When the software system generates a two carload movement from an off-stage shipper that are headed to an on-stage destination, upon arrival one of them has to wait to be spotted until the next operating session. I guess that the customers on my test line aren't in much of a rush to unload their deliveries. This adds to the puzzle that the switch crews have to solve.

CP, AP

CP is constructively placed and I have only ever seen it used on loads.  AP is actually placed.

@Nobody of Importance - The computer won't be solving the puzzle

I agree with you. I wouldn't find it very interesting to just follow the instructions printed out by the machine, especially if they were so detailed that it told me what switching moves to make in what sequence.

The software system that I'm testing at the moment doesn't do that. About the only thing that it does 100% free of human involvement is to generate car orders, but even these are based on criteria specified by a human. It assists in the filling of car orders, the assignment of cars to certain jobs, and in the loading and unloading of cars. Aside from that, it can produce a series of reports using information in it's database, such as switch lists, wheel reports, yard reports, etc...

I'm working on a post that I hope to have up soon that will go into more detail on how this works.

@Dave - shuffling decks of car cards

While I understand that all logistical systems require a certain amount of paperwork, and that sometimes that paperwork because disorganized, I don't think that we should purposely build that event into simulation. I think that the simulation should attempt to minimize such events.

The QR or barcode technology doesn't have anything to do with car forwarding, but rather it helps to make the interface between the humans and the system more efficient and less prone to problems. One thing that we've found in the development of data processing systems is that they tend to propagate human errors faster than paper and pencil systems do.

Yes, computerizing a poor

Yes, computerizing a poor process, or one open to errors, just makes making errors more efficient.

Where I'm torn, is that it's nice if a computer program handles all of the car assignments, etc. But then it also needs to know where all the cars are at any given point in time. In my current system, the computer doesn't know what cars are there, just that there is a car there. The paperwork is with the car (or the agent) to tell you everything that you'll need to know. And if it's missing, then you do the same thing the prototype did, which is find the paperwork. Finding the paperwork consists of me pulling a valid waybill from my drawer of waybills, and putting it in an official railroad envelope and "delivering" it to the agent, who now has something to work with.

So I'll continue to follow this with great interest to see if the system works for me, or if not as a whole, what I can learn from it.

Randy