So we read that Pawn Sensing might be getting deprecated, and that the new system we should use is AI Perception.
Right now we have a simple ai system that uses Pawn Sensing, should we change it to AI Perception?
We don’t seem to find much documentation about it, except
and
And how does EQS tie into this?
Also, I found this
So watching it!
Hey Marcos,
Since Pawn Sensing’s use is so wide-spread, and “new” AI perception is lacking both on documentation and BP-API interface side, we’re not going to deprecate Pawn Sensing anytime soon, especially that there’s no one currently working on ironing out AI Perception’s shortcomings (it’s just BP and docs, mind you! 
).
Having said that I’d strongly encourage you to try AI Perception out due to enormous performance difference, and overall flexibility. Of course if you just need a simple perception, and don’t see it being a perf problem, then you should be fine by sticking to Pawn Sensing. Just bear in mind we’re no longer developing Pawn Sensing and it won’t get new features, ever (bwahaha! 
)
Rest assured, we’re not going take the rug from under your feet Even if we deprecate Pawn Sensing (no sooner then in 9 months if you ask me) it is going to stay around for some more time, and we will show you how to easily switch over from one system to another.
Cheers,
–mieszko
Yeah, that’s what we thought. Since we are at the start of implementing AI we might go with the AI Perception then, is just harder without the documentation but might pay off in the long term.
But regarding a comment I added above, how does EQS tie into this?
It’s kind a separate question, but the answer is short - it doesn’t Regardless whether you use Pawn Sensing or AI Perception, you’ll (just) need to implement custom EQS “contexts”, like “All Known Enemies” or “Current Enemy”. If you need to know more please create a separate thread.