Видео

Wanted to Behavior Trees to develop my game's AI, but my thought process is still stuck at using FSMs, this tutorial is very useful!

What is the software package you used for this video which has the virtual environment and the real-time BT in the sidebar?

At 24:00 you introduce a small heater behavior tree but I think there's a mistake in the condition. It should be T<T0 for the Heater to turn ON and not T>T0.

If actions have to plan around pre conditions and post conditions, It ruins any modularity. You can't just use a generic FreePathToObjectExists() behavior or a generic MoveToObject() behavior... You would need custom variants, like SafeFreePathToObjectAndGoalExists() and SafeMoveToObject(), so those actions don't make plans that cause oscillations by breaking earlier rules. Sometimes, you can't really break down the problem into parts that only focus on individual tasks, some problems require planning the entire route. You need A* pathfinding that labels hazard areas, and it would need to understand how holding an object effects the size of the robot, and it would need to understand how each action effects the paths to all parts of the plan. if Move to object needs to know the entire behavior tree, what is the point of a behavior tree? you will just have to hardcode MoveToObject every time you change the tree.

you explain very well can you do tutorial on Utility AI as well?

I was having problems designing long-running actions. Should they check for conditions themselves to continuing to run? For example, the "eat" subtree checks each tick and fails if there is no more food. But seeming this video, makes sense that the condition checks should take place outside, in the sequence instead, so that "eat" does not execute again if there is no more food. Is that right?

Great video! Is this project open-sourced?

Hi Professor, thank you for the great talk. I have a question: If behaviour trees have essential cyclomatic complexity of 1, and in your book section 2.2.3 you prove that a FSM can be rewritten into a BT, does that mean that FSMs also have an ECC of 1 as well? Additionally, I noticed that that the BT in section 2.2.3 there is an additional global state variable storing the current state. Does having a black board or global variables somehow break this ECC=1 proof for BTs?

2.7K views and no comments, wow! This helped me immensely. It's very well put together, great work!

26:00 magick

Amazing explanation. Thanks!

The volume is low and it ruins the video. Please take care of this issue.

Extraordinary explanation.

First of all, great series of videos, you describe complex designs in a great and very understandable way. Thank you! I believe that you should clarify in the video that you talk about "Deterministic" FSM vs BT. I would like to know your opinion about non-deterministic FSM vs BT. Which are the pros and cons in that comparison?

Great Video! Is there an algorithmic or simple way to check these bugs for large behaviour trees?

At 24:16 are we to understand Eq (5) as the union of all Kbar_j's s.t. the condition holds? Otherwise it is a set of subsets of U.

Damn!

Very informative videos, Thank you!

Thank you very much for the series of videos, it is extremely helpful! However, I noticed that only the first part has subtitles, the other videos do not have subtitles, is it convenient for you to add this? thanks again!.😀

Thank you, very clear compared to other explanations I've found. The book was even more handy. I have noticed a couple similarities between different algorithms - pre- and post-conditions, priorities (utility functions). Any future plans for a BT vs GOAP video address these? I also love how you can nest different algorithms, eg BT / GOAP on FSM, or a purpose built a* pathfinding library on a BT.

Hey Sir @Petter How to connect with you?

Great video. Thanks for sharing

Thank you for sharing!It means a lot to me.

awesome talk, thank you ❤❤

Thank you for the video. This theory introduced in the video is very interesting, but I have a question. Even if an infinite loop is detected in the BT analysis, it does not necessarily mean that the BT is problematic, am I right?

nice!

The video mentioned, "If during that search a person came by a person nearby was suddenly returned success and the agent would stop searching and ask the person if the person could please open the door.". I'm not sure how this could be implemented using py_trees. I think this involves parallel processing, interruption. Could you please elaborate on this?

And how selector choose best Action depending on or ? Or cost

Great video, thanks. But I have a confusion. In the last example (the burglar one), the solution to avoid the contradiction problem is adding a "Hide Camera" action to the right. But in this case, as long as the action of "turn light off" action is made, the rightmost "hide camera" action will still not be taken? Doesn't that mean the contradiction continues? Was the "Turn Light Off" action supposed to be deleted when adding "Hide camera"? Is there something I am missing?

The fact that you are giving out this knowledge for free is simply amazing. This is pure gold, thank you so much!!

Thanks, I have a few questions if you can help understand better. Q1) You indicated that the eat banana was running and this state is returned to the fallback (selector) ancestor. During the subsequent execution of the tree where does the execution start? Does it start at eating the banana (running state) or does it restart with the left branch (eat sandwich / eat apple) all over again or somewhere else? Q2) If the subsequent tree execution starts at eating the banana, how would one allow for interruptions for eating the banana since it now starts at eating the banana? Wouldn't you have to process before this to catch interruptions (ooops turns out banana is bad inside or spouse calls you from the other room) ? Q3) Is the execution of the tree always completed in a single loop of the main sequence of the program? Let's say this is in Unity engine, would the tree execution completed in a single Update? Thanks for video!

Darwin never said "survival of the fittest". Also, if you want to do crossover on things like FSMs, try using the NEAT algo. If it works for NNs, it should work with any other tree structure, don't ya think? :)

Thank you Petter, it was very insightful

Thank you.

Wonderful thanks!

İs This Unreal Engine Behavior Tree?

İs This Unreal Engine Behavior Tree?

The comparison and criticism about RL are just amazing. Looking forward to seeing the improvement of RL methods with behavior trees.

Thank you very much for the series of videos, it is extremely helpful! I have a remark, though: I do not really understand the concept of "RUNNING" actions. If the ticking is synchronous, the action won't return anything before it's finished, right? At that point it will know whether the action succeeded or failed. And if the action is asynchronous and return RUNNING because it's basically doing work in another thread, that means we can't as easily rely on the tree for interruptions because we need to stop the thread or at least alert it. To take the examples from the video: if I'm eating the banana and the alarm suddenly rings, I won't instantly drop the banana and start leaving the building. I'll have some kind of "stop" effect that will let me cancel/pause the "Eat banana" action. Doesn't this basically bloats the complexity immensely in real-life behavior trees? I feel like I really am missing something, there. Does the tree need to "remember" what task is currently running so it can tick it again next time? That would prevent interruptions from working at all but it really feels like interruptions are not as straightforward as what the video says when we try to think about the practice.

Thank you very much for those videos, they are very helpful!

Were at the age where CompSci lectures feature phrases like "Kiting Strategy" :)

Could we implement some kind of parallel container that would act as a « memory » that would record each time we check for a condition both its result and it’s time of verification ? We could then use this memory when a simple belief is needed and actually check it when this is really important to have fresh data. Moreover, when only the belief is needed, we could have a subtree checking quickly how recently has this checked occurred and, if « good enough », trust it as such, if not, check it For the example you used about the closed door before going to bed, if I remember closing the door 15 minutes ago, even though it’s very important, I’m still confident enough that the door is indeed closed, if I haven’t checked the door for the whole day, I might want to check it

It kind of sounds like BTs are the perfectly refactored form of an optimal FSMs…

Thanks for the easy explanation 👍

hi, i'm little late to the series, I'm just wondering what happens if the time taken to compute the success, failure or running status in a leaf is more, what will happen if it can't able to maintain the frequency of the ticks?

ᑭяỖｍｏ𝓼𝐦 ❤️

Oh men this is the best tutorial ever!!!

great material!!!!

I started watching your video without realizing you were the actual author of the book. Thanks for putting your time into this. It might be a bit much to ask for but I would highly appreciate an explanation of the section in your book on stochastic behavior trees. I have studied stochastic processes in the past, but it's still a bit dense for me. Thanks for the video.

Thanks for the video. It's really helpful!