| 
|
Resources
Books
Software
(top)
AI
in Games
(top)
|
"The AI is completely new in SoF2, and
isn't anything like the AI in the
first game. The new LICH AI system makes enemies take cover, lean
around
corners, crouch, rappel, jump, work in pairs, work in squads, leap
over
objects, trip alarms, search for blood trails and bodies etc. and
all with
little or no scripting. It's nice to have the AI independent of
designer
scripting because that makes enemies just as deadly in randomly
generated
maps as they are in the single player game."
- Kenn Hoekstra (Project Administrator for Raven Software)
"The LICH Artificial Intelligence Toolbox
is designed to provide an expandable,
functional, and modular collection of common tools to assist the
AI
programmer. These tools range from low level API modules which access
entity, animation, file, and scripting systems, to Finite State
Machines,
Neural Networks, Path Finders, and much more.
The system is broken down into layers, including
a Game Systems Layer, API,
Motor Control, Senses and Moods, Short Term Decisions, Goals and
Environmental Interaction. The goals of the system are:
- To make enemies use vision and hearing
to detect player
- Enemy AI raises alarms
- The ability to return from full out action to stealth mode
- Enemies will interact with world objects and surroundings
- Enemies will react quickly to player and changes in surroundings
- Enemies will work in tandem
- Enemies will be interested in self-preservation
- AI will have perpetually raising difficulty levels
The AI in Soldier 2 goes far beyond the scope
of anything seen in the
original game. The LICH system overall will make the enemies more
interactive with the environment by default, rather than having
to be told
specifically what to do in each instance with designer scripting
or event
triggers." - Kenn
Hoekstra (Project Administrator for Raven Software)
Related Links
Interview
with Kenn Hoekstra (elited.net)
Interview
with Kenn Hoekstra (pcshooter.com)
|
|
"We retained the functionality for things
NOLF did well, like AIs finding cover and dodging from enemy fire,
and wrapped it in new code for Goals that let AIs make their own
decisions based on their observations of the current state of the
world.
Rather than scripting AIs to enter specific
states, AIs monitor visual and audio stimulus and their proximity
to SmartObjects to determine the most appropriate Goal to activate.
AIs can choose to activate any Goal in their
GoalSet, and assigning different GoalSets to different characters
can create a wide variety of behaviors. The result is more open-ended
gameplay and more variety for the player."
- Jeff Orkin (NOLF2 AI Engineer)
Related Links
NOLF
2 the complete developer interview (pcgamer.com)
No
One Lives Forever 2 AI Q & A (ign.com)
|
|
Uses neural-networks to control the AI racing
cars.
Related Links
Interview
with CMR2.0 AI Programmer (generation5.org)
|
|
Neural networks power Creature Labs virtual
creatures, endowing them with the ability to adapt to their environment
in real-time, learning from experience and forming generalised concepts.
Underpinning each virtual creature is a sophisticated
genetics system. Creature Labs' genetic systems encode complex organisms
with un-scripted behaviour; everything from internal systems such
as brain structure and metabolism through to appearance. Genetics
systems allow virtual organisms to breed, enabling the creation
of unique and unpredicted virtual organisms through competition
and evolution. Genetic exchange and mutation allow virtual organisms
to develop throughout their life and to evolve and adapt to their
environment in ways that could not have been hand engineered or
even envisaged.
Related Links
Creature
Labs Technology Overview
Interview
with Steve Grand (generation5.org)
"Creation:
Life and How to Make It" - Steve Grand's Book
Interview
with Toby Simpson (gamasutra.com)
Creatures
Developer Network
|
|
"The game boasts a new level of artificial
intelligence.
Your Creature is almost a living, breathing thing.
He learns, remembers and makes connections. His huge range of abilities
and decisions is born of a ground-breakingly powerful and complex
AI system.
The villagers in Eden also use revolutionary
AI. Each has his or her own personality, and this can be altered
by what they see, what they do and what sort of person they are."
- About Black & White, Lionhead
Studios
"Creatures learn in a variety of ways.
They learn how to perform actions:
they learn to fish from watching the villagers fishing, and learn
how to
perform miracles from watching the gods casting miracles. They learn
what
sorts of situations should trigger what sorts of desires: you can
train a
creature to be playful whenever he is bored, or train him not to
be
aggressive after he has been damaged. They also learn what sorts
of objects
are most appropriate for satisfying which sorts of desires. "
- Richard Evans (Black &
White AI Programmer)
Related Links
The
Use of AI Techniques in Black & White - by Richard Evans
AI
in Games: A Personal View - by Richard Evans (gameai.com)
Interview
with B&W AI Programmer Richard Evans (planetblackandwhite.com)
Black
& White Postmortem (gamasutra.com)
|
|
"I was in the market in Hong Kong and
decided to lift some
first aid supplies from a stall there. A Military Policeman saw
me do it,
immediately set off an alarm calling for backup, then decided to
enforce the
death penalty on me. I took off running through the narrow market
alleys,
and the MP's were following the parallel lanes attempting to head
me off.
They closed off the road ahead of me and started firing while another
group
closed in from the rear. I ducked through a small china shop to
slip out the
opposite door, when the plate glass windows and vases began exploding
all
around me (just like a John Woo film). A military security robot
had covered
the rear of the store and I just got past him and out into the open.
As I
raced past the gates of an old compound, a stray bullet from the
robot hit a
guard stationed there. The guard was a member of a militant triad,
headquartered in the compound I was racing past. A second later
the gate
swung open, the injured guard returned a few shots and ran inside.
His
cohorts rushed out to see who was attacking the compound and a brawl
erupted
between the gang members and police. When the smoke cleared, the
area was
littered with corpses and one inactive military robot. I got away
with the
Medkits." - Steve Powers
(Deus Ex level designer)
Related Links
Deus
Ex Interview (ign.com)
Interview
with AI Programmer Scott Martin (planetdeusex.com)
|
|
"There are several levels that would
be considered 'AI' in some form or another:
- At the lowest level, there are terrain analysis pieces and pathfinding...
determining the shape of impassable
obstructions, finding
routes around them, dealing with changes as the game goes on.
- Then, there's entity AI. The basic actions that a unit takes,
like gathering, attacking, fleeing, and so on.
- There's also a higher-level terrain analysis to support the AI
scripts. It breaks the map into small pieces of similar terrain
and
resources, and tracks information like "safety" and "resources
available"
for them.
Those parts are all in C++. There's
also a huge "knowledge base", which
works like a giant SQL database of all game information. That's
also in C++.
The scripts (AI and map scripts) run on top of all that. They're
written in
XS, our own custom AI language. It uses and/or directs the lower
level
subsystems. An easy way to think about it is that most of the things
(other
than unit micromanagement) that a human player does are handled
by the
scripts." - Mike Kidd (AI
Designer)
Related Links
Interview
with AI Designer Mike Kidd (planetageofmythology.com)
|
|
"The AI makes decisions on many different
levels. A leader AI keeps track of
high-level concepts, like whether enough science is being generated
or which
civ is the biggest threat. A city AI decides which units to build
for that
area of the world and how to manage the city itself. The unit AI
is itself
split into multiple levels, first determining an ideal destination
and then
calculating the best possible path.
Furthermore, the different AI levels communicate
between each other. For
example, the unit AI controlling a tank might ask the leader AI
which enemy
city would be best to attack. After receiving a response, the unit
AI would
determine its own best path to the destination. Thus, the AI is
able to
organize a large group of units while keeping the discrete game
decisions at
a lower level." - Soren
Johnson (AI Programmer)
Related Links
Interview
with AI Programmer Soren Johnson (apolyton.net) |
Sites
(top)
|
