KISS

I'm a fan of keeping things simple.  RMC ran an article a few issues back about operations, where it described how the local agent would assign the next available freight car to whatever customer needed a car at the time.  That's how I plan on operating my layout.  Customer X needs a box car, so the next available box car gets billed to go there.  When loaded, it goes back.  I suppose I have an advantage on my layout, since I have the yard on one end.  Everything goes in and out of the same yard, which keeps it pretty basic.  I don't worry about where it's going after it leaves my layout, or what the car is loaded with, or anything else that doesn't directly impact the way the car is handled.

As for train schedules, I model a short line where trains were run as needed.  I plan on doing much the same thing.  I do have some schedules (modeling a prototype is a handy thing here), but it's basically one train out and back every day.  I'll run that one train, with a few extras as needed to meet the customer demand and to keep whatever operators I have on hand busy.

More than switching

While my layout has plenty of industries to switch, there's much more to the operating scheme than car cards vs. switch lists (I use car cards).  Whatever method you choose for car routing is probably fine, provided it's clear and concise.  One enhancement I am adding is a simulation of customer demand by instructing crews which of the available cars are actually ready to spot (e.g. the customer has maybe five cars on their way, but only space for four right now, so the switch crew needs to know which ones are to be spotted).  I'm representing this using an idea picked up from Lee Nicholas' Utah Colorado Western, by inserting separate "spot" cards into the car cards.  The waybill may read "To: Customer XYZ" but won't get spotted until the OK is received by the agent.

Beyond that, however, what TRAIN does what is a big part of developing operations for me.  I have a variety of locals and through trains, and most of the through runs also stop somewhere en route to perform a block swap.  I'm still refining this as full formal sessions are approaching.  Each train has specific procedures and fulfills a unique role in the railroad doing its daily business.  For example, a particular eastbound might take cars blocked for the UP to hand over to UP via the Ogden, UT gateway.  Another might only take eastbounds for the D&RGW.  The yards have to keep things blocked to ensure efficient handling.

Operation

We have just had our 4th operating session on the Pennsylvania Midland. Most of the PMRR focuses on loading hoppers with coal at one of two of the online tipples. New Island Coal tipple #3 is served by the Shenandoah & Tidewater Railroad and are taken to the B&O Interchange where they are then hauled either to the Ohio River at Parkersburg, WV or to Baltimore for export.

Cars loaded at New Island #1 are hauled in unit trains of 20 cars to the Allegheny Power plant at Kingston, WV. The cars at Kingston are unloaded inside the Allegheny Power plant thawing shed and hustled back to the tipple for reloading. The entire round trip including loading, travel over the PMRR mainline, unloading at Kingston and return to the mine is about and hour.

All trains from the tipples are run as extras. Those that are handed off to the B&O are emptied by hand in the hidden mole staging that serves as both ends of the railroad. The PMRR track plan can be seen here:   http://www.rockrail.org/pennsylvania-midland

I wanted to avoid the car card/weighbill scenario and have a shipper driven scheme instead. Using the car request generator from Dallas Model Works ( http://www.dallasmodelworks.com/interchange/myindcarreqgen.asp ) I generate switchlists. These are then copied and pasted into my own PMRR switchlist form and given to whomever is going to switch the branch line industries at the several towns. Cars are brought back to the B&O and are picked up by either a through freight or a turn from either Parkersburg or Cumberland. So far it's working well. We'll continue to experiment and refine as we go along.

Operation

For car forwarding I use car cards and waybills.  I wrote my own MS Access program to generate car cards and waybills.  The waybills are single shipment and many are swapped out after completing a car cycle.

Train operation is timetable and train order using a 4:1 fast clock.  One session covers 12 hours of fast clock time and I run 6 through freights, 2 locals and 4 passenger trains.  In addition there are 2 switch engines, one primarily handling yard duties and the other an industry job.

The era is set at 1900-1905 Philadelphia and Reading Wilmington & Northern Branch.

I am also working on 1/4 size waybills instead of CC&WB and changing the fast clock to 3:1 and going to an 8 hour session. 

My take on DB operations

I model the German Federal Railroad (DB) which has a high amount of passenger traffic. For that operation, I'm using a time table based on the actual timetable for the area I model. I've had to slightly tweak times since I don't have the space to model at a true 1:160.  Actual operations of the trains are a combination of automatic running and manual operation by signal indication.

I've also been interested in freight switching and over the years have played with many different systems including switch lists, car cards, and computerized options. I've decided on using car cards based on those shown on  Werner Falkenbach's site, but modified slightly.  At first I was looking at the  Fremo style cards with waybills, but decided that N scale was just to small for reading reporting marks, especially if I weather cars.  There is also a lot of information on the standard Fremo card that doesn't apply to N scale and isn't needed on a home layout.  By having a card that only identifies the car type, any car of that type in the train can be used, making switching a little easier as well. The cards are in German to keep a touch of authenticity. I added a picture of the car to the card. I'd like to be able to operate with friends and family and not everyone will be familiar with the standard Wagengattung (sort of like AAR) letters.  The cards are laid out in Excel and formated so each one fits on a standard business card. I went with a vertical format so the can be read while in the card boxes at stations. They are double sided allowing each car to travel between two destinations. I printed them out using Avery's business card stock at first, but more recently just on card stock and trimmed with a paper cuter.

Each station (two so far) has card boxes from Micromark, one box for each industry or freight house. You can see them here:  http://web.me.com/tommyl/BD_Kaiserslautern/Blog/Entries/2011/3/20_This_and_that.html

​More recently I've been looking a coping prototype switch lists. There has been a lot of discussion about this on some of the German forums that sparked my interest. 

Keeps me way too busy sometimes........

Thanks

Thanks for sharing guys!  I always find it interesting to hear how others in the hobby put there layouts to work.  And to add my own operations to the list, I use something similar to Norfolk Southern's practice. 

For my local switch jobs I use a "work order" which is a single sheet of paper that lists all online customers with cars they have released and cars they have requested for spot.  I use a simple switchlist to keep track of what cars are in the train.  

For my Road Trains I simply just use a switchlist.  

This method is simple to use and very easy to set up.  Car routing is listed in my Timetable and I have built in dwell times for all my industries which simulates the cars being loaded or unloaded.  After the dwell time expires the car is ready for movement for its next destination.

Has anyone tried any computer

Has anyone tried any computer programs like Ship It?  I have heard both good and bad comments on those type of products.  The most repeated comments was that it is complicated to set up and use.

Computer Issues

One headache I've seen from the computerized switchlist generators is what happens when a car gets misrouted or left behind.  The software assumes the car moved everywhere it was supposed to during the session, and any errors or omissions in car routing start a cascade of invalid entries on the switchlists.

Another problem is how many of the programs concentrate so much on individual car routings they ignore train makeup (i.e. the software worries about getting a particular car in a train instead of trying to get cars into blocks and then blocks into trains).  Too often this results in a yard crew having to unrealistically cherry-pick cars to build trains (partially so they avoid the consequences of missing a car routing as noted above).  I think a major part of the issue lies with developers understanding computer programming but not how a railroad runs.

Computer issues are prototypical

That sounds exactly like the way NS's Centralized Yard Office seems to work!  It only works as long as everything is entered into the computer exactly correctly and as soon as it is done - which never happens.  Add to that the fact that the people running the programs do not really understand the people running the trains, and it's a recipe for mass confusion.

I'd rather do things the old fashioned way, on paper.

Switchlist

I get around these issues by ignoring the "specific car to a specific industry" and instead fill the industry car requests with whatever appropriate type car is available in the staging yard(s). If Adams Tool needs an 40' boxcar, the request can be filled with C&NW 164787 or B&O 987764. When the car(s) are delivered, the crew picks up whatever cars are already spotted at the industry if any and returns them to the yard.

It Depends

A lot depends on the size and format of the layout and the number of cars involved.  For a more traditional, over the road layout, involving multiple trains and towns I prefer the traditional car card system.  For switching layouts such as mine, I prefer switch lists or their 'work order' counter part.

One thing I've learned with modern switching operations is the role the customer plays in car spotting.  Often the crew will have a cut of cars carded for XYZ industry but with no specific spotting direction given beyond that.  Upon arriving at the site the conductor and customer rep. will meet and the customer will tell the crew where the cars go.

Lance

Visit Miami's Downtown Spur at http://www.lancemindheim.com

Running an industrial switching district in N Scale

I'm building a what's basically an industrial district with a small yard, an intermodal terminal and somthing in excess of twenty-five possible spots for cars at the industries .  The scenerio is that it operates on both sides of a busy main and is serviced by three scheduled freights a day - one turn from the division and two intermodal trains dropping off and picking blocks of cars for one and two specific destinations, respectively.

My intent is to use a simple spreadsheet that uses the random function (and the number of cars a week expected to be arriving or to be loaded for delivery off-scene) to determine traffic.  On a rare occasion, that may exceed the limits of the A/D track and it'll mean an "extra" turn [not to mention an extra long operating session].  But, whether I run the session as a one man show for my own entertainment or with a second switching crew, what's left at sidings from the previous session becomes the outbound traffic and what the random list determines will be inbound should give me a variety of situations to deal with.

The only wrinkle is when the random function generates more than one car for a sopt and one has to held "overnight" (until the next session) for delivery.

I plan to set up the turn before starting a session (according to the arrival list) and assume any car on a siding from the previous session is ready to be picked up, whether it's supposed to be loaded for a destination or emptied and returnable for reuse.

One side note.... running in N scale where the car markings aren't all that easy to read (and, to be honest, I have  instances of multiple cars with the same markings), inbound cars will have some really ugly tags made from Post-It notes that will give their destination (industry and, where applicable, spot).

I'll probably get more sophisticated when the industrial district is incorporated in a larger layout, but I doubt that, operating in N scale, I'll get rid of the ugly tags (except during photo ops, of course).

It's my version of KISS.  Ask me in six months if it worked....  Well, ask me in a year.....  a year and a half? 

Random Spreadsheet

Our club uses something like that to generate our car orders as well. We use a traditional car card and waybill forwarding operation; the spreadsheet is used to pull the new waybills for staging the next session. These waybills are assigned to empty cars, and those cars are assigned to trains based on their destination. This keeps our traffic mix somewhat variable and realistic.

Basically each car type or pool has a min # of cars, max # of cars and frequency. By tweaking these base numbers, any particular car pool can be as consistent or as wildly varying as desired.

To really refine our operations, waybills are assigned based on detailed pools rather than simple car types. The various pools vary in specificity, but are generally based on car type, assignment, region, etc. This allows some pretty realistic routings. The surest sign that this works is that at any given moment, our major modelled yard is predominantly home-road cars (and various tank cars) with very few foreign railway cars (this is accurate for the particular location and operations we're modelling) and the through trains have a realistic makeup of home-road and foreign bridge traffic, especially the one pair of trains that can be instantly identified from the New England traffic on them.

Computer

I have had mixed results with computer programs.  I have tried 4 different computer systems, only 2 of which are still active.  The only one that generated anything close to satisfactory reports was JMRI Operations.  But, as with all of them you had to adapt to the computer's way of doing things.  JMRI actually generated lists, but I decided that the learning curve wasn't worth the results based on the way my operations worked.  

If you are looking at a list system, even though it didn't meet my needs, I would still recommend it to try  since its free and will generate lists that will meet many operator's needs.

Matching the car movement method to your circumstances

One thing to consider is matching the way you determine car movements to your circumstances and the skill level or involvement level of the crews.

If you are running sessions with others who are not very well-versed in switching specific cars to specific locations, a manually-created wheel report system ("drop off two boxcars at this industry, pick up one covered hopper at that industry") may be the easiest and most comprehensible way to run things at first. With wheel reports, you don't end up chasing car numbers and it keeps things simple.

Car cars and waybills would be the next step up in complexity, by requiring specific cars to be delivered to specific industries. Movements can be determined by shuffling a deck of waybills and drawing as many bills as you want to move cars. If cars are spotted wrong or not at all, you have someone re-spot or deliver the car and card to correct it, or pull the waybill to abandon the move.

Manually-generated switch lists are a bit more work to create each time, but can be just as flexible as car cards/waybills, without the setup time to actually create the car cards and waybills themselves.

From what I've seen on a number of layouts, computer-generated switch lists require more of a focused commitment by the crews running the layout if they are to really work well. They're not as forgiving of missed set-outs as car cards or switch lists. This is especially true of computer-generated list systems that are designed to work in real time, as opposed to on a session-by-session basis.

We decided to use car cards and waybills on the Operations Road Show layout (Come run with us in Grand Rapids this summer!) because it provided a fast and comparatively way to quickly set up work for our crews at each session, even if the initial effort to create the car cards and waybills themselves was pretty big.

We treat the system as being able to self-correct if crews mis-spot cars or forget to set them out. This is important when some of the crews are new to both having to spot specific cars AND run under timetable and train orders. We understand that people will make mistakes, especially since the vast majority of our crews have never run on our layout before, so we've designed our operation in such a way as to easily correct for mistakes.

And that makes it all a lot more fun for us and for our guest crews.

operations without customers?

Not that I'm a big freight fan, but it seems to me that all operations schemes only run the "trains" half of the system.

All those industries carefully modeled on the layout and no-one ever seems to be assigned to manage any of them. Presumably all your car assignments should come from the requests of the "industries" ordering departments for pick-ups and deliveries. And there be some healthy shouting into the phones from irate customers, if cars ordered are not arriving, and/or departing when they should. In real life many of those indutrial businesses could have major disruptions and financial disasters if cars are badly delayed or mis-routed.

Modern Railroading and Computer Systems

In general modern railroads operate off of "Switch lists" and not waybills.  A railroad manager really needs to know one thing....where do I take the car next. 

If I were modeling an older rail operation I would go with the waybill scenario as that is the way they operated.  Physical waybills were passed around.  In today's world the waybill looks like this:

ISA*04*SWAYB     *00*          *02*CSXT          
*02*RMXXGNRR       *120527*1758*U*00501*000539307*
0*P*> ®                                           
GS*WB*SBD*GNRR*20120527*1758*539307*X*006020®    
ST*417*393070001®                                
BX*00*R*PP*089821320001*UP*L*B*S*G®              
BNX*A**S®                                        
N9*BM*089821320001**20120522*1450®               
N9*SI*089821320001®                              
N9*CO*5006226800001**20120523*0948®              
N7*UTLX*666034*187900*N*71200******RR****5505*M® 
N8*518352*20120522®                              
F9*52866*FREEPORT*TX®         

As you can see from the above, the waybill has been translated to an electronic message (EDI) and passed around in this fashion.   However, when the rail manager looks at a track/train he will see something like this actual worklist from the GNRR in Marietta, GA:

 1 LEHX   7000 L  C   AP LEHIGH    BAL                 4427/LEHIGH   
 2 CEFX  82466 L  C   AP LEHIGH    BAL                 4427/LEHIGH   
 3 LEHX   7015 L  C   AP LEHIGH    BAL                 4427/LEHIGH   
 4 BRIX  97292 E  C   PU CSXT      ELZ                 4543/PILPRIDE 
 5 BRIX  97241 E  C   PU CSXT      ELZ                 4543/PILPRIDE 
 6 BRIX  97218 E  C   PU CSXT      ELZ                 4543/PILPRIDE 
 7 BRIX  97398 E  C   PU CSXT      ELZ                 4543/PILPRIDE 
 8 BRIX  97205 E  C   PU CSXT      ELZ                 4543/PILPRIDE 
 9 BRIX  97270 L  C   AP PILPRIDE  CAN                 4543/PILPRIDE 
10 BRIX  97325 L  C   AP PILPRIDE  CAN                 4543/PILPRIDE 
11 BRIX  97237 L  C   AP PILPRIDE  CAN                 4543/PILPRIDE

In modeling a modern shortline (such as the simple one you describe and the one I am modeling) I would stick to creating simple switchlists using excel.  They're easy to make and you can print them quickly and easily.  And cutting and pasting between lists is also easily done.

Customer driven

Good point Andy. Customers are the reason for the railroad. I'm attempting to use Dallas Model Works to generate Car Requests that are then cut and pasted into Switchlists. At present, the program at Dallas includes car requests for all industries including those that have 0 (zero) requests. This is less than ideal since I can't just copy and paste the entire list but rather have to parse through it.

Car requests

You can create a car request generator in Excel or other spreadsheet program pretty simply.

Open a spreadsheet.  Create a tab for each day.  In the first couple columns put the the industry/track  name and location.  Then put a column for each car type you use.  Add a column for comments.  Then duplicate the car type columns to the right of your list (demand columns). To the right of that put a variance column.

In each car type cell for each industry type in a formula to vary the number of cars slightly, lets say plus or minus 20% of  normal.  So lets say the column for boxcars is D, the demand column for boxcars is M, and the variance column is S and Acme Industries is on row 10.  You are operating on a Tuesday, so on Tuesday's tab, you have entered in the demand column that Acme can use 3 boxcars and has a variance of .2 (20%).

In the boxcar cell for Acme Ind. type the following:

=M10*((1-S10)+(2*RAND()*S10))

That will give you a demand between 2 and 4 cars with an average of 3.  By making the variance larger the number of cars will be more random.  By making the demand higher or lower the average number of cars will go up or down.  For example if you wanted to sometime not spot a car you could make the demand 2 and variance 1.  That would give you between 0 and 3 cars.

By dragging the formula across a row of cells it can be copied quickly, then dragging the row of cells down the column will populate a whole table with the formulas.  In an hour or two you can create a customized car demand generator for your railroad.

Customer Demand Generation

Our spreadsheet is somewhat similar, but we explicitly specify a minimum and maximum number rather than an average and variance, and also a frequency (%). This ends up using a couple of formulas as the frequency is calculated first to determine whether or not to even pull any waybills.

The Min/Max works basically the same as Dave H.'s system above. Using frequency as well allows us to have some shipments happen less often. For example, you might want an industry to infrequently get a shipment, but every time they do get a shipment, it's 6-8 cars.

To simulate longer unloading times at certain industries, some of our waybills have a note on them "Hold 2 Days for (un)loading" before the waybill gets turned for pickup.

Already did that

Already did that - and made it available to others as a Google doc.  See:  https://forum.mrhmag.com/post/anyone-want-a-free-switch-list-generator-for-switching-layouts-12189351 for original and some improvements added by others.