Goal-Oriented Action Planning - Advanced AI For Games (Udemy.com)

Another great course from Penny. I've been looking forward to taking this all year, as it's the last course on my professional development plan. I worked with multi-agent system in grad school for my thesis work, so I was excited to take this course. I noticed some similarities in the systems. I've played Condemned: Criminal Origins, but I hadn't realized that the AI in there were part of a GOAP system; now I may want to play it again. I thought the concepts in this course were well explained, though I'm going to go back & digest some of the planning code a little closer to better understand how it is executing the theories explained at the start of the course. I also enjoyed some of the extra Unity game dev tricks, and I'll definitely be putting some of those to use in my own games & other courses.

Really excellent course on GOAP! I learned a lot and am much closer to being able to execute a project I have been planning. This course will teach you how to implement goal oriented action planning that includes agents, world and individual states, and dynamic and static resources. Honestly a very smooth course, the only thing that hung me up on occasion was variable naming made it a little more challenging to grasp some of the code. I was able to get through it, but I would definitely recommend the instructor consider more descriptive variable names for the sections of code dealing directly with GOAP functionality, because it does make it more challenging to grasp the concepts when the variable naming isn't clear.

Great class. Really good to have her refactor code to make it more usable. Also, the environment creation which is the second half of the class is really valuable. It really helps put things in place and provides a great starting point for creating dynamic environments. The only issue I had is that the GOAP system, as it is implemented, feels very tightly bound to objects/people so pre-planning would be extremely important to use it as it is designed in this class. I guess that is an opportunity for me to create a version that better fits my expectations. That said, the instructor does explain how the code works so implementing changes should be fairly straightforward.

This is a great course. You learn by coding (from the ground up) a functioning hospital simulation with different agents (nurses, patients etc) with GOAP AI. There was a lot of copying code upfront, and I was worried that I wasn't absorbing the concepts fully - there's quite a lot to take in. However by the midway point things become clear through the various exercises and sim enhancements Penny teaches through the course. GOAP is fascinating, and Penny is a very clear instructor. Only wrinkle is that sometimes I worry about the number of public classes, and calls to methods like GetComponent() which might end up being hard to maintain or bad for performance, but worrying about that is clearly outside the scope of the course (and I applaud the focus). Great stuff- highly recommended. Thanks! Updating this to say I really valued the extra work to make the test cases generic- really excellent education. Thanks again!

Excellent course! Introduces a great structure for simulating complicated behaviours of characters, allowing you to easily add to the possible actions your characters can take. You'll learn how to give your characters multiple goals and thorough an algorithm select a preferred one that's currently possible, and build out a plan to achieve it. The course includes getting characters to places, allowing them to interact and exchange resources, and interact with the environment. It also teaches game-like elements, like adding resources to the environment via a UI, repositioning and rotating them, and deleting them, as well as some nice camera movement controls. So you can have a nice basis for building out a simulation/strategy type game! The instructions are clear and easy to follow, and introduced in a fun order! The concepts can get a bit confusing, and you may encounter annoying bugs, but just look ahead on the video if so, as the instructor often explains how to solve them right at the time you have them. By the end you'll have a basic Goal Oriented Action Planning (GOAP) system built out, 4 agent types each with a couple of sub-goals, and 4 resource types created and used in different ways. A few more examples of complex chains of behaviours, to demonstrate what can really be done with GOAP (or where it may have limitations?), would be welcome. But you'll learn the basics and have the know-how to figure out how to break down any complex behaviours you want!

Interesting approach to AI. A little annoying tons of refactoring the code, instead of "doing it right" the first time. The AI is working nicely, and i like the packages to download, that allow me to keep updated strictly with what Penny got on her screen. Still im getting a feeling that it's lacking more depth about some more complex behaviours, instead of moving the camera. Complex behaviours i mean: making decisions like do A or do B or do C depending on what is more optimal or simpler in this particular situation. (Fire a gun, OR toss a grenade?) Here the AI is barely waiting for actions to be available and that's all.

Just finished, phew! There's a lot here and while I think I went into it expecting a little more theory (there's an AI course for that!), you will come out the other end knowing how to use the GOAP system. Some Unity idioms that are used later in the course could have been leveraged earlier on to simplify the experience (for example, ScritableObjects make a good candidate for the Actions), but that just means there's plenty of scope to refine and enhance the system to suit the developer's goals. As always, Penny and the team have been active in their feedback which is always reassuring. And wow, once up and running, the system is something that's truly excellent to see; the end results are a huge payoff. Thanks!

This was my first Udemy course so I wasn't sure what to expect. As a c# developer and someone who has used Unity for several years I was worried about purchasing a course that focused too much on how to use the game engine or do basic things. I've on occasion came across some of Holistic3d videos on youtube and they have always been very good. Considering this was an Intermediate / Advanced course I decided to give it a go. I found it very easy to follow and appreciated the instructors use of moments through the video where she gives you a personal challenge to try on your own to reinforce what she just showed you. The final product that you build along with the videos is feature rich and easily extensible. I plan to take more courses soon!

Allow me to say how hugely important this is for a person who is interested in certain type of game design. Creating games where you can create AI behaviours that supports emergent situations. As a mediocre coder i have been trying to learn GOAP for ages. I have read documents, research papers, watched theoretical videos and have gotten a pretty good idea of the theory behind the system but i was still unable to create this system on my own. I am a total "monkey see monkey do" type of learner. Someone has to show me how to build it first. Sadly i was unable to find any good tutorials that would show me how to create a basic GOAP system step by step. But now my troubles are over. Discovered this course a week ago. Everything was explained pretty well and it took a bit of fiddelling around until things clicked. I completed the course and had a basic building blocks in order to continue my own research. The most important is that i finally have a codebase that can i understand. I can take it apart and put it back together, modify and experiment with it as much as i want. I have a fantastic base to finally contunue learning the possibilites of GOAP. This course helped me to get over that annoying "starting hump" where i was stuck for embarrassingly long time. And may i say how fun it is to work with GOAP, experiment with different actions, balancing the values and then let the simulation run and see what happens. Today when working on my game i replaced some of my behaviour tree AI with GOAP and i already saw some interesting behaviours that were less predictable and seemed more natural. Of course it all depends on the game you are building but GOAP AI suits my game needs perfectly. Thanks Penny. Really... thanks. A lot.

This really was a terrific course. I had some doubts going into it and the early parts did have a lot of "copy the code this way" without too much explanation about why we were doing it that way. I'm still not sure I'm confident I understand how the Planner precisely works, but I am more confident going back on reviewing it and even trying to do it myself having gone through and completed the whole course. Once, some of the basic engines of the GOAP system are put in, Penny puts us through many different exercises using it, which serves both to start building the core of a game/simulation and also helps us get intimately familiar with the system we just coded. Things really pick up around Section 5, and the course just gets absolutely fantastic. I especially love all the great, amazing details she puts into everything, really going above and beyond expectations to make sure that we understand things that are going on and that we get results that are pleasing. After explaining a lot of the basics, she actually goes back to some parts and adds some polish or explains alternative ways things can be done. There's a whole section toward the end after we build and work with the GOAP system that's really all about just polishing up the simulation and making sure we have something that is testable and pleasing to work with and this covers some stuff that people tend to have a lot of trouble with such as how to properly rotate stuff, situations that can arise when importing models, setting up a UI, etc. She didn't have to do that, but you really get the sense she wants to help you build something great and just teach you as much stuff as she can. I feel really lucky to have her courses as a resource and it makes me pretty confident about buying anything from her.