Video: Valve's system for creating AI-driven dynamic dialog. Mike Hanson's Blog - A "Dark Intelligence" - Part One. A "Dark Intelligence" - Part One. The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community. The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company. Hello again, So I'd just like to start by conceding that scripted dialogue is not my strongpoint. I tried to make a few films back in my student days and they were, let's face it, pretty hammy... pretty bad!
With that in mind I've been writing, re-writing, then RE-re-writing dialogue for my game, PowerUp. Sure it's all an obvious mishmash of my sci-fi influences, piled up high with every "invasion" cliche in the book... but when PowerUp's HATI (your Holographic Artificial Tactical Intelligence) piped up that there was a "Dark Intelligence" at the heart of our reptillian enemy, it began to dawn on me that I might have a great subject for my next blog. Let's find out if I was right, eh. PowerUp - The Evolution of Enemies Yep. A Place To Begin So there I was. 1. Bot Colony: The AI sim that's 11 years in the making.
Just over a decade ago, Eugene Joseph had a vision. The North Side CEO wanted to see a simulation with AI so clever and in-depth, that it would be able to fully understand spoken English, and respond intelligently. Video games have long relied on dialog trees and canned dialog lines written into the story. Joseph's dream was to create a simulation where users could use their own words to control the outcome of various scenarios.
Joseph and his Montreal-based team spent around four years researching the possibilities, before settling on the idea of pursuing this concept in video game form. Bot Colony was born, and the team spent the next couple of years feverishly piecing together the world's first intelligent-response video game. A prototype for Bot Colony was released at GDC in 2009 -- and this is where we pick the story up. "When the first Bot Colony prototype was shown at GDC in March 2009, we had no idea how far we were from the goal," Joseph tells me. Richard Myles's Blog - Baddie AI in Lost Outpost. The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company. When it came to adding the baddies to Lost Outpost ( The sequel to Outpost:Haven ) we knew that due to the law of sequels ( "Bigger, Better, More" ) the original code just wasn't going to be able to cope with what we wanted, so a total re-write was needed that would make our aliens appear to be fairly smart whilst not sucking up all the CPU time. Node Based Pathfinding [ The dirty way ] We knew that A* wasn't going to work, even with splitting up checks to spread the load, it's really expensive and would hurt other areas of the game. Swarm mode. It's not going to end well for our hero Key for us was having as many baddies as possible, so it was time to look at other path finding approaches.
We're lucky in a way that most of our baddies are aliens. This then continues for each node. Cheers, Stanley Handschuh's Blog - AI Pathfinding Ray Casting. The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community. The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.
There are two major types of pathfinding the first is dynamic the second is static. The difference between the two is whether the path is generated in real-time or predefined. In either case the path can be created and held either on the client or the server. One common form of path finding through open or closed spaces uses ray casting. Each colored dot represents a place the player would visit often. Next after all of the rays are cast the system must search out the intersections. As the image shows there are quite a few intersection points between the Grey, Purple and Brown dots. As previously mentioned the goal of this path is to find a route that does not double back in any way. In the example shown in Image 5 the order is different from the visual reference mentioned originally. Matt Klingensmith's Blog - Overview of Motion Planning.
The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community. The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company. I am a robotics researcher and hobbyist game developer. My specialization is in high dimensional motion planning for robot arms. Motion planning is a highly relevant topic for game developers, as it’s needed every time you want to move an AI controlled character from one place to another. In the course of learning game development, I’ve come across many tutorials which discuss motion planning (usually called "Path Finding" in game development literature), but most of these do not go into great detail about what motion planning is all about from a theoretical perspective.
As far as I can tell, most games rarely use any motion planning beyond these big three algorithms: A* Grid Search, Visibility Graphs, and Flow Fields. Algorithm: WALK TO If not at goal: Move toward goal. Algorithm: BUG.