AI trains stopping - track speed shows as zero

[glcoote]

Hi

I have a problem where AI trains grind to a halt in OR. When I use ALT+F9 to takeover the AI train the reason is obvious as the Track Speed shows as 0. It does not appear to be related to the type of train as it has occurred with diesel, steam on passenger and freight. I take over the train, drive to the next speed limit board and return to the Player train and the AI takes off as normal. One common thread may be that the two routes that I have seen this on are both South Australian, AU SA Adelaide, Hills & SE and AU SA Mid North and Barossa. The separate authors make clear in their information that they shared their knowledge, so perhaps this is caused by the way they set up track in these two routes. Anyway, it is a pity to not be able to fully use my two local routes so any ideas on causes/fixes appreciated.

Regards

Mazda

[R. Steele]

I've had the same situation happen on the Blue Comet Route. I always thought it had something to do with signaling script or dat file.
Have not devoted any time to exploring the problem, and signal scripts are definitely not in my toolbox.

[csantucci]

Does something change if you switch to manual and then back to auto?

[glcoote]

Unfortunately not. It looks like this is something that isn't going to be easily fixed...

[Vince]

The F5 HUD, Dispatcher view will show all AI traffic and if stopped, the reason for the stop.
To decode the HUD please refer to the Open Rails Manual.

regards,

[dcarleton]

The AI trains get stopped at semaphore signals that will not clear.

I tried to solve this once, and concluded that something in the signal logic was causing those particular signal types to read the subsequent restrictive and stop signals and somehow apply that condition to signals which should really be clear. Player trains will see restrictive signals when they should be seeing clear signals. This is all in areas protected with semaphore signals. In areas of the same route protected with target signals or three color light signals or position light signals, the signal logic works as expected.

It may not be an OR issue at all, since it may actually be the signal scripts.

[dcarleton]

My previous thread on this subject:

https://www.trainsim.com/vbts/showthread.php?327131-Blue-Comet-Lines-Semaphore-Signals-Malfunctioning-in-OR&p=1916702#post1916702

Also discussed in this thread (starting with post #10):

https://www.trainsim.com/vbts/showthread.php?326082-New-Blue-Comet-Route-Pass-Activity-Uploading

[Vince]

Hi David,

I'd also suggest looking at the Dispatcher Window. . . from the OR Manual Sec 7.5:

The dispatcher window is actually created by pressing <Ctrl+9>. The window is created in a minimized
state, so to display it in front of the OR window you must click on <Alt+Tab> and select the dispatcher
window icon, or click on one of the OR icons in the taskbar.

You may be able to compare what's showing in the Dispatcher HUD and the train drivers view, with the view in the Dispatcher Window which shows signal states along the path.
This could lead to a logical way to determine if there is a signal script problem.
There is a good writeup on signal scripting language in the MSTS Tech Docs that helped me in the past.
I too suspect the scripts.

regards,

[Don6218]

Yes, this is also happening on the RMD-West Route with the AI trains.

With all the knowledge and skill on the OR Team, they should be able to come with a solution to this "Signal - Semaphore - Script" issue.

With more and more people coming on board with Open Rails, this must be solved to maintain interest in these very fine routes such as the Blue Comet & The RMD series. I wish I had some talent to solve this.

Could a straight substitution of the troublesome Semaphores with "Buried" regular signals (which are already in these routes) maybe solve this issue? Just a thought, because the regular Light Signals do not seem to trouble the AI Trains in OR.

My two cents worth! (psst: we in Canada do not have a Cent anymore so it have to be a nickle. those we have! Hahaha!)

[dcarleton]

Yes, the despatcher window - the first thing I looked at. And when we use it to clear a signal, it goes right back to stop!

One thing that somewhat helped was to edit the AI paths to put a double-reverse point before each troublesome signal. From these experiments I concluded that something was causing the troublesome signals to read the states of signals further down the line and erroneously apply them, causing AI to stop. Since the double-reverse points blocked their view of signals further down the line the AI traffic might continue to move, albeit perhaps a bit slower than intended.

[Vince]

Yes, the despatcher window - the first thing I looked at. And when we use it to clear a signal, it goes right back to stop!
..................snip......................

That leads me to believe it's most likely a signal logic error.
If you could attach the routes sigscr.dat file I will have a look at it and hopefully others more qualified than I will also have a look.
regards,

[dcarleton]

I haven't got the route installed currently, but this is from the download.

[dcarleton]

Also, there's places in the route where the signals are real close to each other, like three signals 100 yards apart. Might that be throwing OR off? Perhaps that PLUS some anomaly in the script? I say that because there are other places in the route where the signals work just fine.

[Vince]

Attempting d/l of the zip attachment . . . Could the routes author provide any help?
......................................

I didn't see anything obvious . . . sorry.

regards,

[Don6218]

Attempting d/l of the zip attachment . . . Could the routes author provide any help?
......................................

I didn't see anything obvious . . . sorry.

regards,

That was/is the same issue as before. It does seem to be something in the signal matrix and/or a fault in the way we expect the Semaphores to play nice with the Searchlight Signals. The one I was having the issue with is on RMD-West EB Columbian with the "Oncoming 05" AI which never starts and ends the activity at the 3/4 mark, the Player train is slowed to 20 mph then stops at a random signal where it should be Green but is Red and it will not change, nor does the Request to proceed work. When in the Dispatcher Window the signals will also not change to allow the AI movement nor the Player train to proceed any further than Bearmouth. The AI Starts in Garrison Yard.