AI.implant from Presagis is a multi-platform artificial intelligence (AI) authoring and runtime software solution. Highly scalable, AI.implant is ideal for populating serious games, simulations, and training applications with intelligent populations of computer controlled entities, including people and vehicles.
Highly scalable, AI.implant is ideal for populating serious games, simulations, and training applications with intelligent populations of computer controlled characters, including people and vehicles, AI.implant enables these simulated characters to make sophisticated context specific decisions and to move in a realistic fashion within their environment.
There are two main aspects to authoring intelligent agents in AI.implant.
1. The AI.implant tool chain allows users to easily interactively, automatically or at runtime author the AI world including a navigation mesh (agent map) & perception data of the simulated world used by agents to make them aware of their surroundings. This world mark-up is used by the agents for dynamic path-finding, path planning and as important meta-data used by agents for complex decision making such as find cover or hide.
2. The AI.implant IDE allows creation of rules used for decision making. Based on a sense, think, do paradigm; both the environment attribution and agent to agent interactions are used to perform complex adaptable behaviours even within dynamically changing environments.
![]() |
![]() |
The AI.implant tool chain offers a set of visual authoring tools for creating AI agents & logic and the AI world representation
AI.implant automatic navigation mesh generation is both part of the above tool suit and provided as a standalone capability for the creation or modification of navigation meshes both offline and at runtime for dynamic updates.
Containing runtime libraries for Windows applications, the AI.implant SDK makes AI.implant integration easy. The production-proven AI tool and Software Development Kit (SDK) makes it easy to create and control all entities in any complex simulation. Offering virtually instant intelligent populations and highly scalable individual human behaviours, AI.implant ensures the greatest fidelity for immersive simulation.
Krauss-Maffei Wegmann (KMW)
KMW adopts Presagis AI.implant tools & middleware solution to help train both civilian and military personnel for the twenty-first century.
Midway Games
Midway Games has standardized on AI.implant tools and middleware for all of its future game title development. In addition to reducing cost and time, this artificial intelligence solution will help Midway Games meet the demand for greater complexity in games being developed for the PS3 and Xbox 360 platforms.
The central feature of this release is the highly improved realism of road traffic and pedestrian interactions including the following:
In addition, other traffic oriented features added to 5.7 that bring increased realism:
Many other improvements have been done for release 5.7 of AI.implant, including:
For more details, please read the AI.implant 5.7 release notes.
The position at which turning starts for RoadNetworks has been improved. The speed at which characters approach and travel around a curve has been improved with the introduction of the slowIntoTurns attribute. The movement is also more realistic with a wide range of vehicle settings.
A new attribute—maxCurveLength—has been added to the WayPoint Network. This defines the maximum distance between WayPoints in a curve, which improves speed adjustments in curves defined by many waypoints.
A number of defects have been fixed in this new release (see the release notes for more information)
Performance improvements have been added:
A person character represents humans in your AI environment. By default, a person is setup with dimension and dynamic values to generate realistic behaviours for a human. Through the AI.DE interface, users can modify the motionType and maxSpeed attributes so this person can be easily customized while remaining as realistic as possible.
A wheeled object represents ground wheeled vehicles in your AI environment. By default, a wheeled object is setup with dimensions and dynamics values to generate realistic behaviours for a ground wheeled vehicle. Through the AI.DE interface, users can still modify the motionType and maxSpeed attributes so this object can be easily customized while remaining as realistic as possible.
AI.implant users will now be able to create a more semantically rich wander behavior. By exposing more high level parameters the new Wander behaviour does more than the former WanderAround behaviour. It allows users to specify the environment in which the agent has to wander and within which environmental constraints it has to evolve.
See the release notes for more information.
Roads/sidewalks/crosswalks may be attributed for contextual pathfinding i.e. only walk using sidewalks and crosswalks. Dynamic areas (mine field/radiated area) or obstructions (people/vehicles/moving debris) influence choices entities make as they automatically plan routes within their surroundings.
Cinematic motion quality & minimal runtime cost. Pathfinding adapts dynamically to changes at runtime for ultimate realism.
perform complex dynamic actions such as jumping over or climbing on top of dynamic debris/climbing up ladders/crawling/fast roping/traversing moving platforms or entering and exiting moving vehicles all while pathfinding
Represent the navigation network at multiple dynamic levels of detail at runtime.
Provides hooks to leverage the host engine's streaming capability and works with hierarchical pathfinding for large scale simulation.
Using AI.implant, brains can be authoring using an easy to use object/parameter based rules, finite state machines, hierarchical finite state machines, LUA scripting, Unreal Script, C++…
Quick and intuitive methods of improving runtime performance and debugging AI logic.
As part of Aeria, AI.implant integrates with STAGE, TerraVista or Creator
AI.implant is also integrated with Autodesk Maya and 3dsmax integrations
In addition to the existing out-of-the-box integrations, Middleware maked it easy to create specific integrations and product customizations