I will second this, navigate is trash. I have been using "TH R route" and the autodrive command so the train goes where I tell it to since the route is preset and the autodrive command does little to no thinking.
'Navigate to' is a rubbish command anyway and shouldn't be used. Like John I had problems with it back in TS12 and I've never used it since.
I am not convinced "Navigate to" is the culprit here. Technically it did work, it got the train where it was directed to go navigation-wise. Also, hadn't had this problem since TS12 until SP4, which I attribute to meticulous use of track direction markers and end of line assets. It is the AI that blundered here - it had 3 choices and it picks one of the two that has the train in reverse for several kilometers, and no longer picks the one requiring no train reversal.
This behavior is nothing new. I had issues with similar setups a decade ago with TS12. I put track marks in the middle of the crossover tracks to ensure the AI has a path. In general, beginning, middle, end works for places like this as it prevents the AI from thinking too hard and getting lost.
Did you use it in conjunction with track direction markers and with end of line assets installed where applicable? Those do make a difference in the AI's thinking. With those assets placed meticulously, I have not had this issue for the same route since TS12 until SP4. It appears now in SP4 the AI has been refurbished to not care if the train runs in reverse for tens of kilometers, and choses that over no reversal.
But your point is well taken if the AI is now indiscriminate with train travel direction: target the crossover & cutoff trackmarks in the drivers command list.
In the very first post in this thread, the list of fixes -Fix a crash using the "Navigate to" command could this be a by-product of this "fix" ?
I know, I saw that, but I don't think so for two reasons- a) as I said above, the train did end up at the proper destination, and does so no matter which of the possible 3 routes is selected by the AI, and b) because I made a very simplified version with only the northwest cutoff track and the "navigate to" worked in Sp4. The AI toggled the 4 crossover junctions and took the northwest cuttoff. A bad "navigate to" command would have not destine the train correctly always, I would get a "no path to destination".
Do you still have an older TRS19 (SP3 or older) installed with the same route?
There is no "older" route and session, the identical route and session gives different results.
if so can you check all switches are in the exact same default position as in SP4? What happens if you set the switches in route, from West to the Toledo portal direction?
See this blog entry
https://forums.auran.com/trainz/ent...Techniques-and-Progress-on-the-Fostoria-Route The objective is to simulate the real Fostoria without intervention.