Ai & Safety Systems

Discussion in 'TSW General Discussion' started by prash, Jul 8, 2022.

  1. prash

    prash New Member

    Joined:
    Jul 8, 2022
    Messages:
    1
    Likes Received:
    9
    Anyone else find it kind of annoying that AI trains don’t follow safety systems?

    On the Harlem line, AI trains pile up behind each other. If the train infront were to move, the one behind it would proceed to follow, adhering to track speed but not adhering to signal speed, which would hold the following train at a restricted speed until the one in front passed the signal block. The piling up of trains would be less apparent since trains would be more spaced together as they are in reality router login.

    Another example is literally in all German routes where freight trains could pass you at 160kmh. If AI freight trains would adhere to PZB restrictions, they would have a max speed if 100/120kmh depending on the type of train.
     
    Last edited: Aug 20, 2022
    • Like Like x 9
  2. redrev1917

    redrev1917 Well-Known Member

    Joined:
    Jul 22, 2021
    Messages:
    3,644
    Likes Received:
    7,991
    One of my biggest bug bears about DRA is that on ICE services pulling into Neustadt on northbound services I am invariably overtaken by a freight service thundering into the station at over 45km/h just to pull up in an instance at a red signal.
     
  3. tallboy7648

    tallboy7648 Well-Known Member

    Joined:
    Apr 9, 2020
    Messages:
    6,567
    Likes Received:
    10,810
    With the Harlem line, Matt tried to act like that trains stopping inches from each other is realistic when that was just a flat out lie. DTG should've just admited that they couldn't get ai to follow the signal speed limits hence the unrealistic bunch up of trains.

    I wonder why DTG can't get ai trains to use safety systems and follow signal speed limits. I assume it would hurt memory but if that is the case, then how much would it hurt it really and would the addons manager help mitigate that since routes will now use less memory when that is implemented
     

Share This Page