AI Testbed
  • Register

 

 

 

 

 

 

 


 

Get the full release version of the AI Testbed (v1.2) below. Caution, quality and stability are not where they would be ideally, time constraints forced release earlier than would have been ideal. The manual AI is generally very good, the computer trained AI not so much (it tends to just completely ignore the player). There's a 50/50 chance of getting either each game.

The thesis produced from this project is now available here: AI Testbed Final Thesis Available

NOTE: The initial survey period is now complete, you can still submit responses but they wont be analysed in detail. Comments are still likely to be read, but that's not guaranteed.

Files:
AI Testbed 1.2 HOT
Commercial license Date 2014-08-24 Language  English System  Windows File Size 11.65 MB Download 1,145 Download

 The AI Testbed is being developed by Foamy as the final year project for his Robotics degree at university. It is a simplified version of the core Hunter's Moon gameplay with the focus on the wolf pack's strategic AI. While almost every element will be ported over to HM, it is not the only priority; the academic side must also be justified. The aim is to produce a pack hunting AI where each wolf is its own independent entity, with no higher up controller dictating where they go with perfect coordination. The rules that govern each of the wolves will then be adjusted, and trained automatically by having the computer play both against itself, and against players. The hope is that this training will produce really interesting, challenging, and unpredictable AI. As part of the academic side, user studies will need to be done within the next month to gather feedback and data on the performance of the trained AI.

In the Testbed game, you spawn as the orange circle shown above, lost in the middle of an endless and randomly generated forest. There will be a pack of werewolves nearby, and they will be looking for you. You'll have a flintlock pistol and a sword, the gun can be reloaded but that'll take time. The sword will work well enough against maybe one or two wolves, who knows after that. So you either stand and fight, and probably die, or you try to escape. If you kill or scare off all the wolves you win, or if you can get a good distance away; if you die then they win.

Features

Multi-Agent System

The wolf AI is based on the Multi-Agent System (MAS) idea. It could also be considered a swarm AI. Each wolf has its own independent state that changes based on what it knows about the world. If it sees all its pack mates running in a certain direction, it is likely to follow. It can't communicate with the rest of the pack directly, they don't talk to each other, just adjust their behaviour based on what the others are doing, and what they want to do to. There's a lot of variability in behaviour due to a lot of different contributing factors, and probability of reactions.

Genetic Algorithm

The Genetic Algorithm (GA) works with the MAS to tune it; since the MAS is so complicated it's hard for a person to tune well. The GA basically runs through a huge number of different MAS settings to find the best one it can. When the time's up you take the best result it came up with and see how it plays. This alone can bring out some very interesting and unexpected behaviour, so on top of the MAS variability, the wolves should be a challenging opponent.

Procedural Generation

The forest used for the Testbed is procedurally generated. That means you can walk in one direction forever and the world never ends, and never repeats itself. It's very easy to get lost and disorientated!

Footprints

Both the wolves and the player leave footprints where they go. They'll fade slowly over time, but not soon enough to save you if the wolves are on your trail. Blood trails will work in the same way.

Wind and Smell

 

Scent is a big part of wolf hunting, so naturally it plays a big role here too. Scents are dropped in the same way as footprints, except their visibility to the wolves spreads in an arc. The wind speed and direction are very important considerations. If the wolves catch your scent then escape becomes unlikely. It's not all bad though, if you know the direction of the wind, and you know the wolves have your scent, then you can have a good idea of which direction they'll be coming from; ambush is a popular tactic by both sides.

Screenshots:

Hunter's Moon

For details and all the latest news on Hunter's Moon development:

See Here.

AI Testbed

For details and all the latest news on Foamy's AI Testbed development:

See Here.

GOM News

For all the latest news on the studio.

See here.

Foamy's Blog

To see recent news from Foamy, the lead developer for Hunter's Moon and the AI Testbed.

See Here.

Latest News

  • New quick status page

    See it here: Status

    I am really bad at keeping the blog up to date, and my plans have to change fairly frequently due to personal, professional, resource, and industry situation changes. This page is an attempt to have something that is quick enough to update that I am able to do it regularly. It also has very little information, so I don't have to worry about my confidence in prediction, or sharing things that are not ready. Every project that has been started will have an entry here that clearly states my current assessment of its status.

    {jcomments on}

     
  • New plans

    I think it's about time to talk about some longer term plans again. I've posted a bit about the current work on the ranged weapon interaction system, and also a bit on the HoO collaboration, but it's been a while since I said anything relating to our primary game ambitions.

    We have some new plans, but it's nothing drastic. The previous plan was to develop Hunter's Moon (the werewolves) and then using all the tech it used to build a general purpose combat simulator (GPCS, as yet unnamed). That plan heavily depended on having a year or so of solid full time development; probably funded through a kickstarter.

    Read more ...  
  • A sign of things to come

    With the musket and pistol kindly provided by the hearts of oak team all attention now focuses on the ranged weapon interaction system!

    I am super excited about this, and I think everyone will love what I've come up with.

    Time-scales are fuzzy, but hopefully not as long as it took to put the AI in place last year.

     
  • Colaboration with the Hearts of Oak team

    This is a big announcement that should have been made a while ago, but now that the AI is implemented on the combat testbed it can't wait.

    A few months ago I approached the lead developers for the Hearts of Oak project with a proposal for collaboration. Basically to provide our combat engine to them in return for publicity, assets, and the joy of contributing to such an important piece of work.

    A summary of the agreement is as follows:

     

    Read more ...  
  • Combat Testbed Updated to 1.3

    Big update with the addition of an AI bot to play against in single player! This is the first pass on the combat AI, a very important milestone.

    Full changelog:

    1.3 - 2015-01-03
    Added an AI bot for single player
        Added as a variety of the orange arena, in addition to empty, and one with training props as before
    Cosmetic improvements
        Added body tilt/lean during turning
    Fixes
        Made players collide again

    Get it here: Combat Testbed