Approaching Closed off locations by doors, via door ways; Looking for different methodologies/reassurances

[Juicy_Fruit]

Hello, I’m looking to get different perspectives on how to handle the approach to doors for a monster ai which then proceeds to bash down the door;

The help I am seeking is regarding methods, currently I’ve set up pathfinding links between the front and backside of doorways to allow the ai to move into the doorway/door despite an apparent blockage. This solution worked until extra door blockages are within i.e. 2+ door blockages exist within" → [] → [] → closed off room]

This causes the ai to stop computing a path. A solution I came up with was to use larger jumper links (from start to end) since individual links do not facilitate the necessary pathfinding; they do not work hand in hand together to help compute a path, however a single large jumper link satisfies the issue but depending on how many doors exists more jumper cables are needed

however a bigger issue occured, this big link causes the ai to repeatedly go forward /backwards after completing the door bash. so then a new solution needed to be identified–while I came up with a potential solution, its just not ideal. I would need to constantly obtain the dot product of where the ai is facing in respect to each individual door, alternate larger jumper cables where one’s face is disabled and inactive when the ai is not “looking” in its direction once a door bash completes, normal path link will be reinstated and the associated jumper links connected to itself would be disabled permanently,

(lets say its going Left->Right, the right->left jumpers need to be disabled)
these are the diagrams using pathfinding links:

The most recent issue has prompted me to turn for assistance since I don’t think calculating the dotproduct for the ai’s fov often is an optimal solution, but its the best I’ve got theorized at this point and probably will resort to if nothing comes of this

*initial links are the large jumper cables

[Foxxive (Vinny)]

I think this should go in #help-and-feedback:scripting-support since this is something that can only otherwise be solved with scripting, no clue what this is doing in community resources since there’s nothing your post provides that directly helps the community in any way

[Juicy_Fruit]

whoops i thought I entered code review, i’m not sure how I ended up in community resources

[Juicy_Fruit]

I’ve devised further a solution;

• encountering computational error in some sort (the OP issue) of designated path via wandering mode I’ll just skip the current path and look for another.

• if error is encountered in a chase (likely to occur from the time it takes to bash down a door) then go into wander mode and do bullet 1.

• if error is encountered in investigation (which is likely to trigger at end of a bash down door sequence or if a target gets far away some how), settle for wander mode and do bullet 1.

still willing to hear any responses about it