Lanelet map convention for static obstacles #2826
maxime-clem
started this conversation in
Design
Replies: 1 comment
-
Please correct me if I was wrong. I think most static objects are either considered as obstacles or some symbols to identify road borders. In such cases, the specific type does not matter that much since most logic is same. I guess planning tags sounds better at least for now. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I would like to discuss the convention for representing static obstacles in a Lanelet map.
Current Convention
The recommended way to represent such information is by tagging linestrings with the tag
type
. For example,type=curbstone
ortype=guard_rail
would correspond to a static obstacle.See https://github.com/fzi-forschungszentrum-informatik/Lanelet2/blob/master/lanelet2_core/doc/LinestringTagging.md for more details.
Current Use
Currently, the
lane_following
module of thebehavior_path_planner
uses the typeroad_border
in its logic to extend the drivable area (https://github.com/autowarefoundation/autoware.universe/blob/89bc5d26c6483172e4041f95266a8c5b91669b33/planning/behavior_path_planner/src/scene_module/lane_following/lane_following_module.cpp#L146-L149).
Issue
Currently, a single
type
value (road_border
) is used as static obstacle. If a map contains other static obstacles (e.g.,guard_rail
), they will not be considered by that algorithm.Dicussion Point
How should static obstacles be represented in the Lanelet map ?
I can think of 2 options:
type=guard_rail
. Planning algorithms should consider a list of tags (e.g.,guard_rail
,wall
,fence
, ...) relevant to their algorithm.type=obstacle
and all its relevant obstacles need to be tagged accordingly in the Lanelet map.Please share other potential use cases or suggestions.
Beta Was this translation helpful? Give feedback.
All reactions