Using Apex Utility AI with another pathfinding

  • September 18, 2017 at 10:56 #21984

    sorry for the, may be, dumb question, but I am very new to gamedev.
    I am working on a naval tactical/simulation game and i am considering using your Utility AI for the AI controlled ships, airplanes, aso…
    Is it possible and practicable to use your Utility AI with another Pathfinding solution then Apex Path/Steering?

    September 18, 2017 at 12:02 #21986

    Hi Funkeyes,

    Thanks for your post and welcome to the world of game development!

    Apex Utility AI does not mandate the use of any other assets – our own included. You can use any pathfinding or steering solution together with Apex Utility AI. Apex Utility AI is simply a decision-making asset.

    You may want to take a look at our RTS Demo, which uses Unity’s NavMesh for navigation, rather than Apex Path. The same goes for this version of the modified Survival Shooter demo.

    • This reply was modified 1 year, 1 month ago by Rami. Reason: Fixed RTS Demo Link
    September 18, 2017 at 12:16 #21989

    Hi Rami,
    thanks for your quick response!
    That sounds great. I think you pasted the wrong link for the RTS Demo link. But i know what you mean!

    As a intermediate C# programmer, what do you estimate for learning time, to be able to design and implement a proper pattern for quite complex strategic/tactical ai?

    Just asking, because I am considering starting very early in the process with very simple actions and decisions. So i can learning by doing?

    September 18, 2017 at 12:39 #21992

    Sorry about the wrong link to the RTS Demo, I have fixed it now.

    Frankly, asking me to estimate how long it will take you to implement something that you call “complex” is an impossible task. I have no idea how long that might take.

    I can only speak for myself, and in my experience it is extremely fast to get up and running with the Apex Utility AI, because you can basically implement one step at a time. I can write a relatively complex AI in about one day, if I have the “game API” ready (methods for querying state and carrying out actions). More complex AIs that include coordination, teamwork, planning, etc. typically take 2-3 days to write and get to a working state.

    What I have found to actually take most of the time in many cases, is the subsequent tweaking process. Typically when I write Utility AIs, I expose a number of variables to the inspector, which allows me quite a bit of tweaking opportunities. The initial setup ‘works’, but getting the AI to work in a great way takes testing and tweaking, which can be time consuming.

    I hope this is an adequate answer to your question, even if it may not be exactly what you were asking for.

    September 18, 2017 at 12:53 #21994

    Hi :)
    I did not ask how long it will take me to implement, I was asking how long it will take me to learn the system, to be able to make proper design and implementation decisions.
    Englisch is not my native language, so probably I sound kind of wired sometimes.

    But anyway, you did answer my question i guess ;)!
    The learning by doing approach is probably the best approach for me. So finding out how to organize my code, step by step.

    Thank you!

    September 18, 2017 at 13:53 #21996

    Sorry I read it as ‘implemented’ for some reason, but I guess the answer applies to both cases.

    In my opinion, the learning curve for getting to use Apex Utility AI is very low and quick. However, designing great AI is a challenge in itself that has nothing do with which tool is being utilized. It takes experience and time to design and implement great AI.

You must be logged in to reply to this topic.