At the risk of having my pass on "Railwire - A Forum for Modelers" revoked, a bit of a digression this week into the operations side of the C&W.
While designing the C&W I felt I needed to settle on both train management and a car forwarding systems so that I knew how much room to leave for their function. I planned on using Timetable and headset radios for train management though we eventually settled on TT&TO using a dispatcher and written train orders. For car forwarding I settled on a label system because that is what other N-scalers in Vancouver were using and I expected (incorrectly as it turned out) that most of my operators would come from their crews. I don't personally like carrying car cards, didn’t want to make room for the card boxes, and had no intention of renumbering all the cars given the work and the difficulty in reading them in a yard of 13' center-to-center tracks. Labels seemed like a very good solution and worked well during sessions. Operators soon learn to ignore them aesthetically as they are too focused on their orders and getting the right cars to the right spots. They also work absolutely wonderfully in the main yard.
The problems with the labels were primarily in the restaging workload. It became more convenient between sessions for me to leave the labels on the cars and to move the cars to restage them(!) then to take the labels off and put new ones on. However even this required 2-3 hours for the 350 cars we move a session and was entailed quite a bit of mental gymnastics trying to ensure the job balance was manageable and the car movements all made sense.
Things came to a head at Vanrail in September 2011 when I hosted three consecutive days of operations. After these sessions I committed to investigate different car forwarding systems prior to the next Vanrail (Sept 2013).
I thought that ideally I wanted some form or computer generated switch list system as it would allow me to accurately model the input and output volume of cars for each the industry– accuracy which I value. However I just couldn’t get the switchlist system to work for the yard without introducing a lot of overhead like creating intermediate switchlists or very prescriptive switching orders (likely requiring dedicated operators which I did not want at all). So I resigned myself to sticking with the label system.
In November of this year I had a bit of an epiphany on car forwarding while attending Northern Rail In Edmonton. Two events triggered it. The first was operating on Scott Sabo’s n scale layout which uses switch lists generated by JMRI. To make things easier on his yard crew, he lets them put small tags showing the eventual destination on the top of the cars as they enter the yard however he makes them remove the tags on the departing cars so the operators need to use the generated switchlists. This seemed error prone to me (and it was a bit) but more importantly I found myself asking him to just let the yard crew leave the tags in place so I, as a new operator on his layout, could easily see where my cars were going without having to resort to the switchlist and reading car numbers! The second event came courtesy of Scot as well when that evening he presented to the attendees a session on how JMRI’s operations system works. This system takes the cars on the layout and establishes logical destination and prescriptive routing for them but it isn’t (I don’t think) industry-centric … more car-centric.
For those of you who have hung in there this long, here comes my personal epiphany. It occurred to me during Scott’s talk that car forwarding systems actually fulfill three distinct functions: demand generation (establishing logical car requirements and volumes for each industry spot), car assignment (matching demand and supply with available cars) and “marking ” (telling the operators what to do with a specific car). And further there is no need other than convention and possibly convenience to use a single mechanism to perform all three functions. Switchlists have their strengths, as do 4 cycle car cards, as do waybills with local agent/operators, as do labels/tags, etc.
For my objectives and preferences an ideal for the C&W system might be:
1) a computer generated demand system (could be as simple as a manual excel spreadsheet if exact car numbers do not need to be entered and the industry spots in the system are “trued up” after each session to reflect the cars that actually make it to their destinations)
2) the same excel spreadsheet could be used to match the gross number or cars at an industry or in staging with the demand and generate a simplified “routing” list which I would use during restaging the layout
3) then use labels, or better yet Scott’s tags which are easier to take on and off, to mark the cars during restaging
Using a computer to generate the demand and a shot at the assignment takes out the mental gymnastics I have to go through while not being reliant on every demand having to be fulfilled for the system to keep working. Using tags deals with off-spots and miss-spots as they will eventually make their way to the right place. And using tags rather than labels makes it much faster to change the car destination rather than either relabeling or worse, moving cars. The tags could copy the color/mnemonic of the existing label system because it works well; and the crew knows how to use it. Finally my timing for moving to a tag system rather than sticking with labels was also good as I have just finished weathering the fleet so I can avoid using sticky labels on my new “works of art” …it’s much easier to stick a label on an unweathered car!
So I spent a week making up the new tags. The tag colour designates the direction/subdivision and if the car is to be spotted on the layout rather than routed to staging a printed mnemonic gives the town and industry spot.
A sample of the tags: Evergreen channel, airbrushed, dullcoted then labeled (if needed)
Nelson Yard
Slocan Forest Products lumber loading
About 350 tags are needed. Here are about 1/2 of them (the other 1/2 being on the layout already). I expect the tags and their storage case will make restaging wayyyyy more convenient *cutting restaging time by more than half*.
Thinking about car forwarding as three distinct but inter-related activities might allow the selection of the right system for each individual’s preference, i.e. car cards for demand generation and assignment but switch lists or labels for marking, or computer generated systems for demand generation and car cards for assignment and marking, etc. Each system has its strengths and weaknesses so why be wed to one approach for the whole thing if there are parts of it you don’t like?
Anyway this is the type of stuff I really enjoy about Model Railroading. I am sure this take on car forwarding system probably is not novel but I have thought of writing it up for the Dispatcher’s Panel (OP-SIG’s excellent journal). Comments welcome!
Oh, and on a more "modeling" note in an attempt to stave off expulsion, weathering the whole fleet is nearly done and will be by New Year's!
md
