RideGroups and max onboardtime

[OneGoogus]

Hey

I’m struggling with a school transport use case (300+ passengers) and need advice on the best approach.

1. Exclusivity (Ride Groups):
   I have groups (e.g., school_A and school_B) that cannot share a vehicle. Currently, I’m solving this by clustering them and running VROOM multiple times, but this is slow (5-6 mins) and leads to sub-optimal results where early groups "steal" all vehicle availability.
   Is there a way to define that certain shipments are incompatible with others, so VROOM can optimize everything in one pass without mixing passengers from different schools?

2. Max Onboard Time:
   I need to limit passengers to a maximum of 60/75/90 minutes in the vehicle.
   Currently, I'm manually calculating the pickup time window based on the delivery window minus 60 minutes. However, this doesn't account for the actual route chosen.
   Is there a native way to enforce a "max_onboard_time" for a shipment, or is manual window-shaving the only way?

[jcoupey]

On 1: no, we don't have an "exclusion" mechanism that would prevent different types of person/things being on-board at the same time. I've seen this request come up previously in delivery contexts where regulations forbids to mix certain types of products. You could segment your vehicles to serve various groups using skills but this would not be ideal in the sense that you won't be planning a full route for various groups in a row.

On 2: that would be #703.

[drogers0]

@OneGoogus, An option for the ride groups is to provide your own travel time matrix and encode these constraints in the matrix.
Set travel times between incompatible stops (because they go to different schools) to a large value M where, M >> vehicle time_window.

Example:

stop_1_school_A -> stop_2_school_A = normal
stop_1_school_B -> stop_2_school_B = normal
stop_1_school_A -> stop_1_school_B = M
stop_1_school_B -> stop_1_school_A = M

Note: Going from school to stop should not be M or your vehicle won't be able to be reused (reverse for afternoon).