 |
modl.ai
|
|
modl.ai
|
Use the following descriptions to gain a better understanding of modl.ai bot terminology.
Also known as the Exploratory space.
The exploration space is the environment that your bot experiences. Your bot uses the exploration space variables to determine whether it has reached a new game state.
When you upload your game to the platform, the brain divides your level into explorable sections. Each section is called a Game State. The exploration space is the part of the game state that the bot experiences. The bot uses the exploration space variables to assess whether the game state has changed and whether to continue exploring the current game state or move to a new one.
Game states reset at the intervals you define. This means that the bot could restore and re-explore game states multiple times throughout a test run.
Use the exploration space to divide the level into game states and define how often the game states reset.
Decide how you want your bot to know if it’s reached a new game state and add any relevant variables to the exploration space.
For example, if your character moves around a 2D or 3D space you should add the position variable as the exploration space. This means that the bot will assess whether it’s reached a new game state based on the bot’s position.
You also need to define an Interval (Unity) or Granularity (Unreal). The interval lets the bot know how big a game state is. Choose a small interval if you want the bot to thoroughly explore your game. For example, if you add the character position as an exploration space variable, the interval is based on how far the bot has moved.
Also known as Exploration game state.
The game state variables are what you want to know about. Your bot reports on any interaction it has with your tracked game state variables as well as custom event triggers and performance issues.
When you run a test on the platform, you receive a test report. A test report outlines all of the Game state variables that you decided to track throughout the test run, as well as some custom events and performance information. For example, if your game requires your bot to pick up coins, you can add the coin pickup mechanism to your game state variables to view where and when your bot picked up with the coins.
Note: You must add any variables from your exploration space as game state variables if you want to receive information about them in a test report.
Add any variables that you want to know about to Game state variables (Unity) or Exploration game state (Unreal). To track specific events in your game, use Custom events. Custom events are tracked events that you set up in your code. For more information, refer to Unity’s Custom events guidance or Unreal’s Report events guidance.
A prefab or component in your game that attracts the bot. modl:test uses waypoints to direct the Directed explorer bot through your level.
As your bot moves through game states, it assesses whether it is closer or further away from a waypoint. If it is closer to the waypoint it continues on to the next game state and assesses again. If it is further away, it goes back and tries a different game state. The bot continues exploring as it completes these assessments.
You can add multiple waypoints to your game and the bot uses the waypoint index to decide which waypoint to move toward first. If you assign a waypoint with an index of 1, the bot considers it the highest priority waypoint.
Use the following steps to add and enable waypoints in your game:
ModlWaypoint prefab (Unity) or component (Unreal) to your level.
1.9.1