Does anybody here have trouble with the run around train command?

JonMyrlennBailey

Well-known member
It seems to cause Drive To (track mark) and Move to Train at Track Mark commands on schedules to act stupid.

For example, on my cloned West of Denver route, I ran my my GP-38 engine pair around the passenger train at Union Station in Denver, Colorado. The line terminates there so trains have to go back out the same way they came into the platforms and navigate back to the main line. I used the Change Direction command following the engine runaround to establish the new front of the train. I also made sure the driver was moved to the new lead engine using the Move to Train at track mark commands and these worked there.

This train then drove back to Sulphur Springs via several passenger stops. I used Drive To track mark to stop the trains at platforms at Winter Park, Fraser and Granby. The train stopped at the correct spots at Winter Park and Fraser but oddly stopped way ahead of the platform at Granby. The train oddly stopped so that the arse-end of the train was at the prescribed track marker for this purpose and not so the lead engine was at the track mark as it should have otherwise been.

Then at the staging area at Sulphur, end of route, the driver was supposed to be moved from this train to another passenger train there on a holding track. The Move to Train at track mark failed this time. The driver did not swap trains and the schedule was involuntarily stopped so the Continue Schedule appeared on the drop-down menu. Even when I clicked on Continue Schedule, nothing happened. This Move to Train at track mark command malfunction doesn't seem to occur if there is no run around train operation on the schedule.

I now had to reconfigure my passenger trains so there are engines on either end so the run around operation does not have to be employed at Denver Union Station. Drivers no longer switch passenger trains at the Sulphur holding area to boot. Each train there now has its own driver and trains line up one behind the other on a single holding track and are released one by one every three hours.
 
To the best of my knowledge, AI trains should ideally change direction automatically almost immediately after uncoupling after passing a junction
Perhaps you may have the change direction command setup wrong?

Try this setup Drive to, Wait for, Run Around, Wait for, Change Direction, Wait for and continue the journey
 
I use the run a round command all the time and it works all the time, but then i do not use change train command.
The same driver stays with the same train.
 
To the best of my knowledge, AI trains should ideally change direction automatically almost immediately after uncoupling after passing a junction
Perhaps you may have the change direction command setup wrong?

Try this setup Drive to, Wait for, Run Around, Wait for, Change Direction, Wait for and continue the journey
I have been running some more trials. The train in question still stops incorrectly at the Granby station westbound track marker even with locos at either end of the consist. This was even after Run Around Train was omitted from this schedule altogether. I still moved the driver to the lead loco at Denver Union Station to travel back the opposite way from the platform. I like to have the driver always in the lead engine. Some locos generate a cab figure while others don't. One passenger train I have has a steam locomotive on one end of the train and a diesel on the other. Naturally, if the steam loco is the lead engine at a grade crossing, I don't want to hear a diesel horn sound (HONK!). I don't want to hear a diesel electric at the head of the train blow a steam whistle noise (TOOOO-TOOOOO!) to boot. Having the driver assigned to the cab of the lead engine ensures the proper horn sound for the locomotive kind.

But not all passenger trains are stopping wrong at Granby west! A different passenger train with different engines (Illinois Central E8, <kuid:376853:105513> IC EMD E8A Mid by jacksonbarno) and the Santa Fe Superchief consist is getting this right, I've observed. The train that has been goofing up is my Rocky Mountaineer consist with the following Union Pacific engine content:

<kuid2:338228:300339:3> UP GP38-2 564-605 EX-SP 1 by Adamovic packaged with TS 2022.
 
To the best of my knowledge, AI trains should ideally change direction automatically almost immediately after uncoupling after passing a junction
Perhaps you may have the change direction command setup wrong?

Try this setup Drive to, Wait for, Run Around, Wait for, Change Direction, Wait for and continue the journey
I have been executing Change Direction following the Run Around and always use a 5 second buffer betwixt such commands.
 
I have been running some more trials. The train in question still stops incorrectly at the Granby station westbound track marker even with locos at either end of the consist. This was even after Run Around Train was omitted from this schedule altogether. I still moved the driver to the lead loco at Denver Union Station to travel back the opposite way from the platform. I like to have the driver always in the lead engine. Some locos generate a cab figure while others don't. One passenger train I have has a steam locomotive on one end of the train and a diesel on the other. Naturally, if the steam loco is the lead engine at a grade crossing, I don't want to hear a diesel horn sound (HONK!). I don't want to hear a diesel electric at the head of the train blow a steam whistle noise (TOOOO-TOOOOO!) to boot. Having the driver assigned to the cab of the lead engine ensures the proper horn sound for the locomotive kind.

But not all passenger trains are stopping wrong at Granby west! A different passenger train with different engines (Illinois Central E8, <kuid:376853:105513> IC EMD E8A Mid by jacksonbarno) and the Santa Fe Superchief consist is getting this right, I've observed. The train that has been goofing up is my Rocky Mountaineer consist with the following Union Pacific engine content:

<kuid2:338228:300339:3> UP GP38-2 564-605 EX-SP 1 by Adamovic packaged with TS 2022.

If a train stops incorrectly, as long as you are driving to a track marker and not a station. You can change the default 20 meter distance to around 2 meters. It's found under the advanced tools in the Track Markers toolbox. Change the 20.00 to 2 and click on the "apply" tool and click on the marker in the route you want to change the distance of. That should help all AI trains stop exactly on the track marker
 
If a train stops incorrectly, as long as you are driving to a track marker and not a station. You can change the default 20 meter distance to around 2 meters. It's found under the advanced tools in the Track Markers toolbox. Change the 20.00 to 2 and click on the "apply" tool and click on the marker in the route you want to change the distance of. That should help all AI trains stop exactly on the track marker
Thank you, sir. My test has concluded that all my trains are stopping correctly at this particular track marker except for one. I'm running a new test with different loco content on the maverick train that's been goofing up.
 
Update: I tried different engine content for my Rocky Mountaineer consist and now the train stops correctly at the track marker Granby station platform westbound and I can even use the Run Around Train command again at the Union Station in Denver without such failures as I had before.

I still line my passenger trains up on staging siding outside of Denver. The trouble was that if there was a lead train holding at a track mark at the front of the siding waiting for a timed release onto the main line, trains from behind directed to pull up to this track mark by assigned schedules would wrongfully back out of the siding. I fixed this by rigging up a pair of junctions with a track marker between them and a track marker just behind the lead train waiting. Trains are now lured onto the siding but can't back out of it because the junction pair is set wrong. AI cannot control the lever on the far side of the junction pair with each lever set wrong. This junction pair serves as an anti-reverse mechanism. The track marker between the switches lures the driver forward past the lead switch so he can control the trailing one as well. Trains on the siding now can only go forward and not backward. The switches on the siding act like one-way check valves. When the lead train takes off onto the main line, the following train on the siding (captured by the junction trap) pulls up to this track mark when the previous train moves out of the way. Forward direction is now the only path out of the siding. Yes, track markers are like carrot sticks to coax stubborn mules (or iron horses) forward.
 
I think that far too many users put far too much faith in what is a very basic AI system - after all it is not even at ChatGPT level.

The solutions to most (all??) Trainz AI problems always seem to involve putting in additional infrastructure (invisible signals and markers, carefully placed switches, etc) that would not be practical in a real railroad.

A video that I recently posted in the Rail News forum showed a real life "AI" system running the huge ore trains on 1,000km round trips between the mine-head and the port (with automated loading and unloading at each end). Their "AI" is only used on the mainline. In the marshalling (classification) and service yards human drivers take over.
 
Back
Top