obstacle avoidance

I'm working on a solid solution combining pathfinding and obstacle avoiding.

The first ray of hope you can watch here:

The objects just react on their direct environment eliminating any collision, avoid any type of obstacles and try to find an other way to their goal (in the center).
This way the objects interact "chaotic" what looks a bit like insects.
There is no "path" included now - the objects have just one point to go to (removed when reached).

The next step of programming will be the recognation of "moving" and "static" obstacles and the interaction with moved objetcs.

looks very cool.

they need a some sort of paths. Not exactly paths but like next-point-to reach thing. like you add waypoints to corners, apply a* it finds next point to reach using a*, but moves using the code shown in video, when it reaches that point then it moves to nex point until it reaches it's destianation. optimizing this to to smooth out the movement(i mean when there is no clear visual sign that i goes from point to point, then it's smooth) would make it even better.

then this would be an awesome obstacle-avoided-pathfinding-thing.

i guess there is a lot of c_trace going on there..

Yes - paths to guide the objects...
The first video shows the stresstest for "short-distance" obstacle avoidance (i made it a little harder than ingame expected ).

The framework for it will be my flex_path system:

Here the agents randomly select one node to reach and calculate the shortest path to it. The short distance obstacle avoidance in this version is not finished.

Damned if that ain't the best piece of AI I've seen for GS in six years. Brilliant!! When will it be for sale?

No offence, But have you been living under a rock for the past 6 years?

When I tried intense x it wasn't really that great because you couldn't do anything but have some ai enemies shoot at you. If you wanted to add any of your own functionality you would need to disable the intense x. It was too laborious to use the intense x functions on their own.

It would be better if you had created a library of modular functions which could be incorporated into an existing code base rather than force any user to understand the whole code system before they can do anything original.

As it was when I got it, intense ai was just like a FPS creator clone, which is extremely limiting and boring.

I have been looking at mercuryus pathfinding code and it looks very simple and efficient, the pedestrian avoidance looks just as good as yours. Why it looks better is perhaps because it will actually be usable.

in fact intense-x is good too(i have it,from a contest) downside of it is that you have to rely on the intense-x. If you start with intense-x and develop over it, it is ok. But if you want to add it to a existing project, your project will need a lot of change(since it would be easier to change own project rather than intense-x).this is where the pain comes.

mercuryus' video looks great, looking forward to it. (btw did you tried anything about usin threads for ai calculation?)

No offense taken. I'll have to come out from under my rock more often. I usually only come out on ground hog day to check my shadow.

I'm stuck on A6 and it would cost me $160 to upgrade to A7 and use IntenseX. A little roadbump.

What you see here is available in our Free Intense PathFinding software.

If you are still using A6 not a problem; Intense Pathfinding 1.03

This will be my last post here because its not my intention yo hijack this thread, I only want to point out that you don't have to wait for this pathfinding solution, this is nothing that we haven't been offering for free to this community for Years.



I am sorry, I guess its not for everyone. I can say however we have many users now adding their own functionality easily. I guess some users just prefer to spend a little time learning the inner working of our product than spending thousands of hours to write their own AI code base.

There have been hundreds of Game studio users "Using" our products for years

mercuryus, You product looks great, I just wanted to point out that there are other options available while they wait for you to finish.

Thanks for the A6 link.

I like the top (first) video the most... more than those agents. I think it looks cooler and more fun, it's more abstract, the agents seem more like an attempt at realism that isn't quite finished yet. But you said they weren't finished, so it's fine. their movement is just like... like I'd never move

again, very nice work

I'm happy you like it!

The both methods will be combined.
Think about a basic node related pathfinding plus the obstacle avoidance when obstacles accour.


I'm still working on the obstacle avoidance.
Also i have to find a solution for "Controlling Agents in High-Density Crowd Simulation"

Video not from me:


PS: to behave the individual "more human" the trick is to set more waypoints and let the individual direct to the next visible node with the shortest way to the destination.
This way the individual will walk on shortcuts - like a human does.

"bi-directional flow" is the coolest part

i like part with coffee
really cool stuff

Similar to people in a crowd:

It would be very interresting to make a simulation of
cars in the city or on a high-way.
By djusting the behaviour of the cars you can show
why traffic jams happen, and
prove/disprove that speeding or slow driving is the cause.

Also the layout of the streets can be planned this way.