College Project

Doomofman

Hey all, I'm in my final year of a Software Systems degree and am deciding my final year project this week. I've come up with an idea that might interested the folks on this forum and it's something that I think the insight of you airsofters may help with.

The Idea:

The idea is to created a sort of automated demonstration of breaching a building and either rescuing a hostage or just eliminating the enemy(and more scenarios will be added down the line once the core system is up and running)

So basically it'll be computer vs computer but you'll be able to change several variables etc every time you run it so it'll never be the exact same twice.

I figured since there is alot of experience with military/police tactics in airsoft that people on here might be able to offer some ideas and tips on the strategy side of things and I hope it'll be interesting too.

Thanks for any input people give

TheDoc

Sounds a fantastic idea

I think its called counter strike

Haha but seriously thats a brilliant idea. I remember in days long gone, we used bots with waypoints to instruct new clan members how we go about playing matches. I assume you need to build something from scratch?

I'm sure there is plenty of people that can help you round here

Doomofman

[DM]-TheDOC- wrote: »

Sounds a fantastic idea

I think its called counter strike

Haha but seriously thats a brilliant idea. I remember in days long gone, we used bots with waypoints to instruct new clan members how we go about playing matches. I assume you need to build something from scratch?

I'm sure there is plenty of people that can help you round here

Yeah it'll be built from scratch... I'm reckoning top down view so like counter Strike 2D if anyones ever seen it.

At the moment I don't plan on the suer being able to control the characters, just change some stuff at the start and then have the AI make the decisions about whether to breach hard or try stealth and stuff like that.

TheDoc

hmm, I've always been interested in these type of things, and as far as I know there is something actually out there at the moment that does something similar.

To be fair, I'm not going to clog up the thread haha, cause I really really cant help at all. I'm doing It myself, but my project for this year, is to have a mobile app to track the movement of buses through gps....that idea came from the lecturer..

If the programme works out though im sure it will get huge itnerest in here, be well helpful to put up skirmish sites maps and run some strats for teams to review over.

best of luck

DisasterIRL

[DM]-TheDOC- wrote: »

To be fair, I'm not going to clog up the thread haha, cause I really really cant help at all. I'm doing It myself, but my project for this year, is to have a mobile app to track the movement of buses through gps....that idea came from the lecturer..



best of luck

This idea has already been thought of as far as I know, students in ITB(my college) created a something similar and is now in use by URBUS

Doomofman

DisasterIRL wrote: »

This idea has already been thought of as far as I know, students in ITB(my college) created a something similar and is now in use by URBUS

Shush you

I don't mind if it's been done before.... Its something I wanna do rather then something I'm just stuck doing...

mcgovern

Doomofman wrote: »

Shush you

I don't mind if it's been done before.... Its something I wanna do rather then something I'm just stuck doing...

And it was done in Trinity many moons ago, and even been trialled by Dublin Bus themselves afaik.
But hey, I'm sure it will be more fun than the one I did!

Doomofman

mcgovern wrote: »

And it was done in Trinity many moons ago, and even been trialled by Dublin Bus themselves afaik.
But hey, I'm sure it will be more fun than the one I did!

Oh lol I thought they meant the tactical simulation thing I'm gonna do

Dar

Make sure you do some research and ensure you're not biting off more than you can chew. I can imagine any state machine handling this sort of behavior becoming very, very complicated very, very quickly . You'd have to model not only individual entities, but also model the team work and coordination between them. Look at the amount of effort that goes into commercial game AIs.

I'm not trying to dampen your enthusiasm, just want make sure you go into it with your eyes open .

Lemming

If this is going to include 3D modelling, I would suggest that you think long and hard about this because it sounds incredibly ambitious for an individual effort in college when you will still have other assignments and subjects to concern yourself with too.

The 3D Modelling aspect alone is well .... challenging enough. DirectX; to use an example of one of the prominent 3D APIs *; has a very steep learning curve and for 3D rendering in general you had better be comfortable with mathematical calculation.

Add in some sort of A.I. and tbh (and speaking from personal experience) you may be setting yourself up for less 'challenge' and more 'very rough ride'.


* API: Application Programming Interface

Doomofman

Lemming wrote: »

If this is going to include 3D modelling, I would suggest that you think long and hard about this because it sounds incredibly ambitious for an individual effort in college when you will still have other assignments and subjects to concern yourself with too.

The 3D Modelling aspect alone is well .... challenging enough. DirectX; to use an example of one of the prominent 3D APIs *; has a very steep learning curve and for 3D rendering in general you had better be comfortable with mathematical calculation.

Add in some sort of A.I. and tbh (and speaking from personal experience) you may be setting yourself up for less 'challenge' and more 'very rough ride'.


* API: Application Programming Interface

TBH I think I'm going to have to leave it 2D because as you said it would definitely be biting off more than I can chew...

I think I can do something good enough to show that given more time I could make something more I guess 'finished' or 'complete'.

Lemming

Doomofman wrote: »

TBH I think I'm going to have to leave it 2D because as you said it would definitely be biting off more than I can chew...

I think I can do something good enough to show that given more time I could make something more I guess 'finished' or 'complete'.

The only thing to bear in mind here is that 2D programming may involve programming inside a 3D model (as strange as that may sound to the uninitiated). Certainly the DirectX API abandoned DirectDraw (2D) and concentrated with Direct3D (3D) for its rendering capabilities with version 8 and that was years ago. You can create 2D but you have to do it through Direct3D obviously.

I would suggest between now and your project proposal, that you do a monkey-load of reading up on whatever rendering API you run with (I don't have much experience with the likes of OpenGL), and their ability to render 2D. I would also seriously consider some alternative proposals as well.

Note: I've ignored the A.I. part of this because I have no idea what form you want your AI to take. Are you going to use fuzzy? genetic algorithm? neural network? a hybrid? etc. That's a whooooooooole other kettle of fish to just the graphical side of things.