matlab/matlab-agentic-toolkit / skills-catalog/automotive/roadrunner-build-rrhd-map
roadrunner-build-rrhd-map
VerifiedBuild, convert, and import RoadRunner HD Maps (.rrhd) in MATLAB. Author lanes, lane boundaries, markings, junctions, signs, barriers, and speed limits with roadrunnerHDMap; convert Lanelet2 (.osm, Autoware) maps to RRHD; resolve RoadRunner road asset paths for markings, crosswalks, signs, signals, and barriers; and import or build .rrhd or OpenDRIVE (.xodr) files into a RoadRunner scene. Use when creating road networks for driving simulation, converting Lanelet2 or other HD map sources to RRHD, looking up road asset paths, or loading and verifying a map in RoadRunner or RoadRunner Scene Builder. Not for building scenes from recorded sensor data, running simulations, or recreating OpenSCENARIO (.xosc) scenarios.
references/lanelet2/extractTopology.md
View on GitHub# Extract Topology
Determines predecessor/successor relationships between lanelets using boundary endpoint node matching.
## Primary Method: Boundary Endpoint Node Matching
Two lanelets are sequential if their boundary ways share endpoint nodes. This captures ~96% of lane connectivity including intersection connections that proximity methods miss.
**Critical: Account for opposing boundary directions.** Left and right boundaries of a lanelet may run in opposite directions (common in bidirectional roads where a center boundary is shared between opposing lanes). The "last node" of a reversed boundary is at the lane START, not end.
### Detecting Backward-Right Boundaries
Use the dot product of left and right boundary directions with a threshold of -0.3 (not 0):
```matlab
lPts = geomMap(ll.leftWayID);
rPts = geomMap(ll.rightWayID);
lDir = lPts(end,:) - lPts(1,:);
rDir = rPts(end,:) - rPts(1,:);
dp = dot(lDir(1:2)/(norm(lDir(1:2))+1e-10), rDir(1:2)/(norm(rDir(1:2))+1e-10));
isRightBackward = (dp < -0.3);
% Inverted lanelets (dp >= -0.3 but left way geometrically on the right) have BOTH boundaries
% Backward: their effective start nodes are the last nodes of both ways. Derive start/end nodes
% from the per-lane alignment (conversionPipeline.md, Invariant 3) rather than from dp alone.
```
When `dp >= -0.3`, boundaries go in the same direction (normal case). When `dp < -0.3`, use **proximity** to determine which direction the right boundary travels relative to the lane:
```matlab
if dp >= -0.3
% Normal case: right boundary goes in travel direction
rightStartNode = rightWay.nodeRefs{1};
rightEndNode = rightWay.nodeRefs{end};
else
% Opposing boundaries: check proximity to determine travel direction
d_ls_re = norm(lPts(1,1:2) - rPts(end,1:2)); % left start to right end
d_ls_rs = norm(lPts(1,1:2) - rPts(1,1:2)); % left start to right start
if d_ls_re < d_ls_rs
% Right boundary is BACKWARD: effective start = last node, end = first
rightStartNode = rightWay.nodeRefs{end};
rightEndNode = rightWay.nodeRefs{1};
else
rightStartNode = rightWay.nodeRefs{1};
rightEndNode = rightWay.nodeRefs{end};
end
end
```
### Building Endpoint Maps
Build `endToLanelet` and `startToLanelet` maps using effective right boundary endpoints:
```matlab
endToLanelet = containers.Map('KeyType','char','ValueType','any');
startToLanelet = containers.Map('KeyType','char','ValueType','any');
% For each lanelet: add to endToLanelet(rightEndNode) and startToLanelet(rightStartNode)
```
### Opposing-Direction Filter (MANDATORY)
After connecting lanes via shared endpoint nodes, filter out false connections between opposing-direction lanes. Use **EFFECTIVE travel direction** (not raw right boundary direction):
```matlab
% Compute effective travel direction for predecessor
predDpLR = dot(predLDir(1:2)/norm(...), predRDir(1:2)/norm(...));
if predDpLR >= -0.3
predTravelDir = predRDir(1:2); % normal: right = travel
else
predTravelDir = -predRDir(1:2); % backward right: reverse
end
% Same for successor...
dpTravel = dot(predTravelDir/norm(...), succTravelDir/norm(...));
if dpTravel < -0.3
continue; % Skip opposing-direction connection
end
```
**Why effective direction?** For backward-right lanes, the raw right boundary direction is OPPOSITE to travel. Using it directly in the filter would incorrectly flag valid same-direction connections as "opposing."
### Left-Gap Filtering (MANDATORY — removes false successors)
After building pred/succ connections, apply the left-gap topology filter to remove false connections caused by shared boundary endpoint nodes. See [references/leftGapFilter.md](../authoring/leftGapFilter.md) for the full algorithm and code.
**Summary:** For each lane with multiple successors/predecessors, compute left boundary gap. If any connection has gap < 3m and others > 3m, remove the far ones. Applies bidirectionally. Typically removes ~34 false connections on the reference 184-lane map.
### Why Not Simple Node Matching?
Simple `wayLastNode == wayFirstNode` without flip detection creates false connections for opposing-boundary lanes:
- Lane_124's right boundary ends at node X (which is at lane START due to reversal)
- Lane_56's right boundary starts at node X (which is at lane START)
- Simple matching says "Lane_124 → Lane_56" but actually both STARTs are at node X
- Results in 71m+ gaps between "connected" lanes
With flip detection, effective end/start are correct and all gaps are either ~0m (direct) or ~3m (intersection crossings).
## Fallback: Center Line Proximity
For unmatched pairs after node matching, use geometry proximity:
```matlab
THRESH = 1.0; % meters
d = norm(endPts(i,:) - startPts(j,:));
if d < THRESH
topo(i).successors(end+1) = lanelets(j).id;
topo(j).predecessors(end+1) = lanelets(i).id;
end
```
## Endpoint Gaps After Topology
Connected lanes will have two types of gaps:
- **~0m**: Directly adjacent lanes sharing a boundary
- **~2-4m**: Intersection crossings (lane center to next lane center across junction)
Do NOT snap intersection-crossing gaps. Only snap gaps < 1.5m. The ~3m gaps are real physical distances — RoadRunner uses topology references for routing, not geometric coincidence.
## Performance Note
The O(n²) comparison is acceptable for maps with < 1000 lanelets. For larger maps, build a node-to-lanelet index for O(n) lookup.
----
Copyright 2026 The MathWorks, Inc.
----
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |