A quick and dirty explanation of CSX CADS screens. Pictured is the computer display for most of the Old Main Line Subdivision that runs between St. Denis and Point of Rocks. Screens are arranged so the the Westernmost point is at the Top of the screen and the Easternmost Point is at the bottom of the screen. (On North-South territories the North is at the top and South is at the Bottom.) Every effort is made so that Control Points and Signals are displayed on directional lines; this is displayed on the ‘Frederick Branch’ that runs between Wimer Street and Frederick Junction. The illustrated S curve is just a graphic device to link Minnick and Frederick Jct together in their proper geographic positions. The Frederick Branch is used mostly by MARC commuter trains, 3 each way each work day. The yard betwwn East Yard and Minnick is the ‘home’ of the MARC Frederick Operations. Originating trains leave East Yard and go to Wimer Street to originate and then operate on the branch to Frederick Jct and then continue off this screen to Doub and ultimately turn onto the Metropolitan Subdivision at Point of Rocks The Blue squares that appear on top of the switches at East Yard and Minnick are Switch Blocks that the Dispatcher has applied to prevent operation over those tracks without the authorization of MARC personnel - Blue Flag Protection.
The ‘To 5301’ notation signifys where this screen connects to a adjoining screen. The 5301 Control Point is known as Doub. Likewise at the bottom of the screen the ‘To 5460’ indicates that screen connects to the 5460 Control Point which is St. Denis. The White ‘blobs’ that are below the Main Track representation indicate Defect Detectors, there are 3 on this screen - Illchester, Henryton and Ridgeville.
The occupancy of trains on the model boards are indicated by RED track segments. The direction of the train is indicated by
Yes! It is one of 4 active screens used by the BC Desk that is responsible for the Capital and Metropolitan Subs in addition to the Old Main Line. Additionally the BC Desk has screens that display adjoining terrritory so that they know what traffic is headed toward them.
One feature that CADS supports is Automatic operation. When in Automatic, the track display lines turn White instead of Tan. Additionally Train ID’s are turned White instead of Tan. Automatic can be implemented by the Dispatcher on specific controlled segments of track as well as individual Train ID’s. ie. Tracks between A & C can be in Automatic but tracks between C & F can be in Manual. For the System to line signals both Track Segments and Train ID’s must be in Automatic. If one or the other is in Manual, the signals must be lined by the Dispatcher.
Most Dispatchers do not use the Automatic function, as way too many that have used it have gotten themselves in various forms of ‘trouble’. The logic that Automatic uses is far from straight forward, especially when dealing with multiple tracks and slow orders. (I have seen Automatic put a train through a 10 MPH crossover to avoid a 25 MPH slow order). To my knowledge, neither CSX or Union Switch & Signal (the OEM for CSX’s CADS system) have made a concerted effort to focus the resources necessary to improve the logic in Automatic to make it viable.
My understanding of the NS meltdown around Chicago several years ago was managements requirement that their Dispatchers us NS’s Automatic function exclusively. I may be wrong - but that is the information I recieved.
CSX in hiring new Dispatcher’s has hired a lot of former military Air Traffic Controllers. These individuals have found Train Dispatching to be more difficult than ATC because they can’t change the field of play or the players, as can be done in ATC by changing flight paths and/or altitudes. In railroading, the tracks you have are all you have and you can’t control the altitude of the trains that operate on those tracks.&n
Quoting Balt: “My understanding of the NS meltdown around Chicago several years ago was managements requirement that their Dispatchers us NS’s Automatic function exclusively. I may be wrong - but that is the information I recieved”
Now I know why I was able to spend two nights instead of just one between Chicago and here \when I came back from a trip then–we left Chicago more than two hours late, and arrive here 7:45 late.
We kept losing time–about five hours late at Lincoln, and because the Lincoln crew had been called for on time work, they had to be relieved somewhere in west Nebraska.
I have a book, the memoirs of a Milwaukee Road/Soo Line train dispatcher, titled, “Trains don’t fly.” The title comes from an incident when the author was training a former Air Traffic Controller. (Some of those fired by Reagan in the 1980s were hired for dispatcher’s positions.) The trainee was said to have told a train to change altitude.
One of the rare instances when less complexity - only 2 dimensions instead of 3 (compare 3-dimensional chess in Star Trek; see alsohttps://en.wikipedia.org/wiki/Three-dimensional_chess ) - makes the intellectual task more difficult, not easier. From an engineering perspective, though, in 2-dimensional train dispatching there’s 1 less ‘degree of freedom’ and 1 more constraint than in ATC. That greatly reduces the number of possible solutions (permutations) to a comprehensible quantity, but also greatly limits the options, which are non-negotiable.
Another factor is the relative traffic density, however that can be quantified. Directing airplanes that are ‘short’ and miles (minutes) apart can be different than long trains or closer together. Probably the best parallel would be be a terminal ATC where the number of possible routes - runways - are limited to 1 or 2, same as with a railroad line. Also, there’s a lot of traffic wanting to use them at the same time, or like 2 - 3 minutes headways - kind of like a commuter rail terminal. In contrast, way out in the country the railroad still has only 1 or 2 tracks to handle movements in both directions, while the ATC has the whole sky to work with.
Then there’s this article about what can happen when the pilots let the systems handle all the de
I see two failures contributing to that accident. One was the failure of the pilots to ask the sector controller to repeat his instructions and were guessing at what they were. The failure to request an altitude change was the other. All pilots learn in basic flight school the altitudes that should be flown based on direction of flight.
The flight director/auto pilot only knows what is programmed into it. It will not make altitude changes on it’s own if it is properly working.