Skip to main content
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [sumo-user] Question regarding E3 detector filtering and nextEdges behavior for overlapping routes
Regarding your questions:

1. yes. nextEdges is the correct mechanism but it must be used very carefully to avoid matching the detour vehicles
2. the detector must be placed on an edge directly ahead of the junction and the nextEdges attribut must hold the id of the first left-turn internal edge (ahead of the internal junction if one exists). See the updated documentation at https://sumo.dlr.de/docs/Simulation/Output/Multi-Entry-Exit_Detectors_(E3).html#attributes
3. the data failure likely occured because you mixed normal and internal edges. This is not supported. Either normal edges or a single internal edge are permitted. In your case the internal edge is necessary (though there is an implementation idea to make this work with normal edges: https://github.com/eclipse-sumo/sumo/issues/18267

regards,
Jakob

Am Mi., 26. Aug. 2026 um 16:35 Uhr schrieb Pedro Oliveira via sumo-user <sumo-user@xxxxxxxxxxx>:
Good morning! The SUMO developers likely have a better answer than the solution I am about to propose. One alternative would be to use E1 detectors (instant induction loops). They will return the IDs of all vehicles that pass over them; you would place one at the entrance, one at the exit, and one at the return point. You can import the XML into Excel (that is the tool I am most familiar with, though there may be other methods). Using a pivot table, you can identify the entry-exit OD pair and filter for those that pass through the return point. Best regards, Pedro Oliveira.

Em qua, 26 de ago de 2026 10:12, 郭松庭 via sumo-user <sumo-user@xxxxxxxxxxx> escreveu:

Dear SUMO Development Team,

I am currently working on a traffic simulation project using Eclipse SUMO and have a question regarding the multi-entry-exit detector (E3) and how to properly filter specific vehicle paths.

Problem Description: At a specific intersection, I have two types of vehicles passing through a left-turn movement:

  1. Green Path (Target): Vehicles that make a direct left turn.

  2. Red Path (Unwanted): Vehicles that pass through the intersection, make a U-turn or detour down the road, and eventually merge into the same exit lanes, thereby triggering the same E3 detector.

image.png



Since both paths share the same entry and exit lanes (or end up at the same E3 exit), the E3 detector counts the red-path vehicles as part of the normal left-turn movement, which distorts my performance data (speed and travel time).

What I have tried: I attempted to use the nextEdges attribute to force the E3 detector to only capture vehicles following the exact sequence of the green path. However, whenever I define nextEdges, the E3 output consistently returns -1 (no data), even though the vehicles' routes in my .rou.xml file contain those exact edges.

My Questions:

  1. Is nextEdges the correct attribute to filter out detouring/U-turning vehicles (the red path) when their final exit lanes are identical?

  2. If so, what is the exact mechanism or syntax required for nextEdges when dealing with OSM-imported networks that consist of multiple sub-edges? Why does it usually result in -1?

  3. Is there a recommended best practice in SUMO to separate or filter out these overlapping/detouring trajectories within an E3 detector without causing data failure?

Any advice or guidance would be greatly appreciated. Thank you for your time and support!

Best regards,

Tony


_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

Back to the top