Dogwood Companion System

a new AI Technology

Designed & Built by
Dogwood Gaming 
with Unreal Engine 5.

Large-scale complex AI generation for the creation of dynamic environments in a 3D space.


Development Overview

The Dogwood Companion System is a new hallmark product from Dogwood Gaming designed from the ground up using Unreal Engine 5. The purpose of the Dogwood Companion System is to provide 3D creators with an easy-to-use tool to create and scale large numbers of AI-controlled characters that can be inserted into a 3D space and provide life and realism within that space without requiring a large amount of processing power on the user end. The Dogwood Companion System has been prototyped and developed for projects within video game development, prototyped and developed for projects within video game development, architectural development, and film with the potential to be implemented into projects within a wide scale of fields and industries.

Ashes of Kanaka

Originally developed for use in Dogwood Gaming’s in-development title Ashes of Kanaka, the Dogwood Companion System (DCS) was first created as a solution to populating large 3D spaces with unique and dynamic background non-player characters. A non-player character is an AI-controlled character in a video game and is traditionally relegated to a set path and minimal interaction with the game world. Using a simulated hierarchy of needs and an array of motivations, Dogwood Gaming has been able to create an AI script that allows for a more realistic characterization of these background characters.

Unreal Engine 5

Using Unreal Engine5’s Blueprint system, Dogwood Gaming has been able to develop an easy-to-access set of values that determine a non-player character’s goal within a space. By combining this data set with objective-oriented markers in the 3D space, the DCS creates both a problem and a solution for each of the actors created within the program. This combination results in an AI-controlled character that can move about a space, interact with objects and characters around them, and react to situations to complete their objectives.

Dogwood Companion System

The Dogwood Companion System generates these blueprints for a character either through randomly generated or user-defined sets of traits on a large scale. DCS can create thousands of these actors simultaneously and populate space with them. The DCS allows a user to select or define a set of characteristics that the system can then pull from will create a realistic and lived-in simulation of populated areas.

Technical Overview

The Dogwood Companion System is designed to be easily modified to fit exactly what the user needs. All variables are exposed and ready to be added to or removed from.
The DCS has a plethora of uses, and this example architecture project’s variables and systems can be used for film, games, and other simulations. The base programming of what Dogwood Gaming has created is easy to bring together projects and integrate into various systems.

Building Data Sets

In this example, the system can be seen in use in tracking specific sets of data for building.

Customizable Values

These nodes can take an average of every variable the client wants to monitor. These default values are average levels taken from preliminary research. The values here can be customized to fit a specific region or type of individual based on where the building is located.

Finding a perfect match

If the AI cannot find a perfect match, it will take into consideration all of its choices and navigate to the space that it would find most comfortable. This level of thinking and decision-making can be used effectively to create a space to appeal to as many different people as possible.

End-User Features

Define a set of characteristics

From an end-user perspective, the Dogwood Companion System can easily and efficiently define a set of characteristics and assign them to a group of AI characters without the need for complex processes to run in the background.

Apply characteristics to a set of characters

The plug-in will allow the user to define and select characteristics from within a hierarchy of Master Categories, then apply them to a set of characters either randomly or through user input.

Create needs and personalities

Once these characteristics are selected, they can be modified within a range to affect the importance of the characteristic to the character, creating a hierarchy of needs and personality for the character.

Replicate and Generate

This can then be replicated within Unreal Engine 5 to generate a large number of these characters within a 3D space.

Industry Use Cases

Learn about how the Dogwood Companion System is currently being used.

In the spring of 2022, Dogwood Gaming entered into an agreement with iNESS360-an architectural technology company focused on data-driven analysis for energy efficiency and cost reduction in commercial development. The DCS has been combined with their technology to develop real-time simulations of human behavior within a building space. These simulations track the preferences of a group of people throughout a building layout and can be used to determine energy consumption needs for lighting, HVAC management, and network stability within a building. iNESS360 and Dogwood Gaming have developed a working prototype of this technology and are moving towards scaling for production and marketing of the technology on a business-to-business level.