Embedded Files
World and Objects Representation
World and Objects Representation
In the AI.Bootcamp Engine, all objects have physics attributes. Most objects are simply using basic colliders, but others are a bit more sophisticated. Here's how each object is represented :
Tiles
Each tile uses a mesh convex collider with 6 anchors attached to the middle of each edge and 1 anchor attached to the center of the tile (in red in the picture). All anchors are used as reference locations to attached objects (like walls) and to determine where to raycast when trying to detect if a tile is visible or not.
Walls
Each wall uses a box collider.
Windows
Each window uses a box collider but requires an extra attribute to let the raycast passthrough. This attribute is implemented by using the Layer system in Unity.
Doors
Doors are a bit more complex than other objects, being composed of 4 components : 2 outter panels and 2 inner panels. When a door opens, the 2 inner panels "slides" into the outter panels, leaving an open space to get through the door. Each panel uses a box collider.
=> Note that side panels will block vision raycasts even if the door is opened!!!
Pressure Plates
Each pressure plate uses a mesh convex collider. Pressure plates are not blocking raycasts to ease the detection of the tiles under them. This is done using the same Layer system used for windows.
NPCs
Each NPC is using a mesh collider matching its cylinder shape. NPCs are not blocking raycasts to ease the detection of the world surrounding them. This is done using the same Layer system used for windows.
Google Sites
Report abuse