Hi, I'm running a scenario that I downloaded from ATS. It fails about half way through with an AI collision. I see a stuck AI train that causes the collision - it should move but does not and a following train hits it. When I run the scenario at 32x speed in the scenario editor, the stuck AI train moves off right on schedule and the collision does not happen. I can delete the stuck train to remove the collision and fix the scenario. Does anybody know why a train behaves one one way in the editor and another when the scenario is running? G
It's just trial and error, adjusting timings and testing. The more AI the more difficult and strange it can get. In the scenario editor preview function your train is ai controlled where in the scenario play you might arrive later or earlier at waypoints, influencing how AI drives. I've had AI trains even ignoring red signals and running into me. By moving them a little and changing Start times, adding stopping destinations I was able to fix that.
I write my own scenarios, and add the AI in batches starting with the furthest away first. Test and make sure it works, and then move onto the next point up the line. Everytime I create a version I save it and call it V1, V2, V3, etc. Until the AI is complete. One particular scenario I have is a modern day Newcastle - York, that I just swap the stock to the train I feel like driving (ie HST, class 800, 91 etc…) and adjust the weather. Currently have 11 versions, all of which the AI works perfectly fine. On the 12th version of the scenario, the 7th AI train won’t leave from just Outside York, and causes AI 8 to crash into it. I have no idea why it won’t depart, even though it worked fine in the previous 11 versions! Spent a few hours trying all the above solution to get it to move but to no avail…. So left it broken and filed the issue under ‘it’ is TS … not me’ line of issues! Moral of the story … your not alone! If you can advise what scenario it is, happy to download it and test it and see if I get the same result.
Hi, Thanks westcastlerail. It's this one, a Javelin run from St. Pancras to Ramsgate. There is a stuck Desiro at Strood. The crash happens about 30 minutes in, about the time the player is scheduled to reach Rochester. It takes 100 minutes end to end so no worries if that is a bit long to try. Note that others have had the crash as well. https://alanthomsonsim.com/downloads/rf-class-395-1f14-st-pancras-stp-to-ramsgate-ram/ I note some comments in the scenario page that I am not the only one with the problem. I am going to swap the Desiro for a 465 and see if it continues to happen. That would tell me if it was a problem with the Desiro or something else.
I don't think they type of train makes a difference - it will either move when it is supposed to or not. Sometimes it can happen if two AI trains are too close together on the same line. The AI train should see a red and stop but if it is too close it does not and hits the train in front. Try moving one further back
Thanks for the link to the effected scenario … unfortunately I don’t have anywhere near enough of the requirements for that scenario. In the comments on the ATS site for the scenario in question, there are comments about someone having problems with getting the Javelin to move. Not sure if they are yours or not, but if not then someone else is having the same problem. They recommend installing the latest version of the route (Danny’s route????) … that might work? Other thing I noticed is it mentions the player having to change from AC to DC Power. I’m not sure whether the AI would be advanced enough to do that if required?
The routes are not electrified at a type of voltage but just electrified - the wires or 3rd rail don't mean anything If you wish to run an AC Electric to Brighton no problem - A Southern Emu on OHLE again no problem On London to Faversham you do notice the Javlin AI running pantographs down
The train did make a difference in this case! When I swapped in a 465, the crash didn't happen because I passed the train in the Strood tunnel. I'm not sure what the problem was with the Desiro from ATS.
I don’t know if post four on this thread might shed any light on the issue? “There's a known issue whereby the 'internal' scenario clock is different to the displayed time. You can see this in the predicted arrival time at the next waypoint or stop, and in the fact that you can sometimes get full 'on-time' points for a stop even though you were slightly late by the displayed time. This may be related. Depending on the question you're asking, a different answer might be that you need to adjust the effort percentage in the scenario instruction - if you want it to predict / require an earlier arrival, increase the effort percentage.” https://steamcommunity.com/app/24010/discussions/3/3199244571764652888/