Interlocking Towers with AI consists

[TheBrook]

I am sorry to raise another interlocking tower issue, but I seem to have a problem with AI controlled consist trains. I have read many posts regarding interlocking towers which have helped me greatly to identify my key issue. To conduct my tests, I have created a simple track with 2 loops with a common side (like a squashed figure eight). One train travels around one loop in one direction with another train traveling in the opposite direction on the 2nd loop, with the common side with 2-way traffic being managed by an Enhanced Interlocking Tower asset. After much trial and error, my conclusions are:

• When in manual driving mode, with individual locomotives or consists, everything works perfectly, paths activated and cancelled as expected.
• In AI mode, when individual locomotives are used, everything works perfectly.
• However, in AI mode, when consists are used, the interlocking tower path is cancelled as soon as the consist passes the ENTRY signal, rather than waiting until it passes the Exit signal. This is shown on the Interlocking Tower Manager runtime monitor. I even downloaded a consist from the DLS to prove there was not something wrong with the way that I created my consists, but it had the same problem.

I use Trainz Plus, on build #123794 and using "Enhanced TRC3 invisible Interlocking Tower", <kuid2:61392:8101:78>.

Can anybody advise if I am doing something wrong?

 

[TheBrook]

I am sorry to raise another interlocking tower issue, but I seem to have a problem with AI controlled consist trains. I have read many posts regarding interlocking towers which have helped me greatly to identify my key issue. To conduct my tests, I have created a simple track with 2 loops with a common side (like a squashed figure eight). One train travels around one loop in one direction with another train traveling in the opposite direction on the 2nd loop, with the common side with 2-way traffic being managed by an Enhanced Interlocking Tower asset. After much trial and error, my conclusions are:

• When in manual driving mode, with individual locomotives or consists, everything works perfectly, paths activated and cancelled as expected.
• In AI mode, when individual locomotives are used, everything works perfectly.
• However, in AI mode, when consists are used, the interlocking tower path is cancelled as soon as the consist passes the ENTRY signal, rather than waiting until it passes the Exit signal. This is shown on the Interlocking Tower Manager runtime monitor. I even downloaded a consist from the DLS to prove there was not something wrong with the way that I created my consists, but it had the same problem.

I use Trainz Plus, on build #123794 and using "Enhanced TRC3 invisible Interlocking Tower", <kuid2:61392:8101:78>.

Can anybody advise if I am doing something wrong?

After further testing, I can further refine the problem scenarios. When in manual driving mode:

• With individual locomotives (not consists), everything operates as expected.
• With consists, the interlocking tower path is cancelled as soon as the consist passes the Entry signal when the route is a closed circuit i.e. it is a loop with no junctions turned against the path of the train.
• With consists, everything operates as expected if at least one junction is turned against the path of the train.

Similarly, in AI mode:

• With individual locomotives (not consists), everything operates as expected.
• With consists, the interlocking tower path is cancelled as soon as the consist passes the Entry signal when the route is a closed circuit i.e a loop, with or without junctions.
• With consists, everything operates as expected if the route is a point-to-point circuit.

I hope this helps identify the problem. Can anybody assist with this issue?

 

[Kennilworth]

Hi


When you're setting the path up you will see that the top line is "Path clear method" which has four options. Experiment with these and see if one of them makes a difference.

I'll often throw together a simple route to test an idea out but I find that the best way is to try it on an actual route. My own preference is to use the option "Clear on drive" which allows each junction to clear as the consist leaves it but this can be changed depending on the circumstances of a particular situation. Loops of track as you describe can be difficult to get to work correctly and need to be signaled carefully. I have made extensive use of EITs to control single line sections over the past 6 or 7 years and have no issues with them.

They do have a steep learning curve which can be frustrating but it is worth persevering with them and I use them on all my sessions now.

Regards

Brian

 

[stagecoach]

I would suggest that the EIT is seeing the rear of the consist. As you drive to the entry signal it triggers the path, but because the EIT can also see the end of the consist it thinks the train has cleared the path. EIT can see a long way down a track and the only way to shorten the distance is to use EITPath triggers.

 

[TheBrook]

Thank you both for taking the time to respond to my issue – much appreciated. I think Stagecoach is spot on when he thinks the EIT is seeing the end of the consist. I subsequently did try the use of EITPath tiggers. They did delay the start of the path selection, which I can see would be useful, but it did not seem to prevent the EIT seeing the end of the consist and thus it still released the path upon passing the entry signal.

Although you both may think it strange that I am playing around with loops, it is because I am creating a simulation of my physical model train layout. Model train layouts often have ‘continuous running’ by means of some sort of loop. I want to see how realistic can a simulation be compared to a physical layout. My physical layout makes extensive use of automated complex operations so I am beginning to think Trainz may not be suitable in that aspect, although it has many other wonderful features.