Scenarios
Scenario JSON files keep map size, terrain profile, obstacle definitions, benchmark seeds, semantic tags, and recommended vehicle settings outside runtime code.
Scenario files
File |
Use case |
Notes |
|---|---|---|
|
Static obstacle navigation |
Baseline map for planning, sensing checks, and deterministic experiments. |
|
Dynamic interaction |
Useful for moving-agent and obstacle-avoidance experiments. |
|
Narrow passage mapping |
Canyon-like traversal and local mapping. |
|
Alternative map profile |
Visual and navigation variation for regression checks. |
Map actions
Action |
Payload |
Result |
|---|---|---|
|
|
Regenerates the map by scenario name. |
|
|
Generates a named or default procedural terrain. |
|
|
Returns dimensions, path points, scenario name, and map metadata. |
|
|
Returns occupancy and class grids for planners. |
Dynamic agents
Dynamic AUV-like agents are managed separately from static obstacle geometry so planners can distinguish fixed map structure from moving targets.
API |
Purpose |
|---|---|
|
Create a dynamic proxy AUV or obstacle agent. |
|
Teleport or reposition an existing dynamic agent. |
|
Drive a proxy agent with an external velocity vector. |
|
Configure underwater communication latency, bandwidth, range, and packet drop rate. |
Agent schema fields
Every scenario declares an agents array. The main AUV entry is metadata for
the physical vehicle already present in the scene; non-main entries can be
spawned by the scenario loader.
Field |
Purpose |
|---|---|
|
Version of the agent definition format. |
|
Stable runtime identifier such as |
|
Logical role such as |
|
|
|
Vehicle profile or runtime profile name. |
|
Initial pose fields. |
|
One of |
|
Sensor names or sensor declaration dictionaries. |
|
Role, semantic label, tags, and experiment-specific annotations. |