Configuration parameters for real-time business rules of pick-drop orders

String Collection: A string separated by semicolons ( ; )

Team Configure parameters

In the real-time pick-drop scenario, the following configuration parameters are involved, starting with the team level:

Parameter Name	Type	Description
enabled_rule_codes	String Collection	Enable those "preliminary selection" rules. Available rules:basic;area_code;tolerance_check;geo_merge;max_pick2drop_merge;worker_radius
meter_search_radius	Integer	How far away can an order be assigned to an employee
meter_search_radius_whitelist	Integer	If a whitelist is enforced, how far away can an order be assigned to an employee?
peak_period_spec
enable_accum_items
enable_level_items
enable_skills
allow_mixed_area
meter_same_area
meter_other_area
enable_accum_items
enable_accum_items
limit_same_location_stack	0
peak_period_spec	String Collection
period_unit_second	3600
meter_same_location_match	10
meter_priority_merge_pick	300
meter_priority_merge_drop	1000
meter_allowed_merge_pick	2000
meter_allowed_merge_drop	4000
meter_merge_drop_scaling	20000
meter_max_pick2drop_merge	1000
tolerance_seconds	0
nbr_jobs_per_slot	0
...	...	...

Rule lists

Based on the above parameter combinations, the pick-drop orders support the following business rules:

Order Dispatch Priority

All orders and riders belong to an area code (area_code). If your business operates in a single area, you can specify a constant such as "A". When a new order arrives:

• Riders within the same area who are within meter_same_area distance and are idle will be prioritized based on the closest proximity.
• Riders from other areas who are within meter_other_area distance and are idle will be considered next, also prioritizing the closest ones.

Configuration can be made to prioritize either order consolidation or selecting idle riders during specific peak periods:

• If order consolidation is prioritized, orders will be consolidated based on pickup within meter_priority_merge_pick meters and dropoff within meter_priority_merge_drop meters, prioritizing distance. Other riders will not be considered.
• If selecting idle riders is prioritized, idle riders within meter_same_area kilometers from the order will be prioritized first, followed by riders from other areas within meter_other_area kilometers, directly assigning them without considering riders already carrying orders.

If the pickup and dropoff addresses for an order are the same, and limit_same_location_stack equals 0, an unlimited number of orders can be stacked, even if they exceed the configured maximum.

If no matches are found through the above priority searches, the algorithm will look for other orders to consolidate based on configured maximum order counts and maximum distances, selecting the allocation that minimizes the "newly added distance" across all eligible riders.

In any case, if consolidation occurs, the distance between the dropoffs of the existing and new orders must be within meter_allowed_merge_drop. The following three parameters jointly control the maximum distance for consolidation:

• "max_meters_radius_merge_drop": "500",
• "max_meters_radius_merge_drop_scaling_distance_km": "15",
• "max_meters_radius_merge_pick": "2000" For example, if the current order distance from restaurant to customer is 5 kilometers, riders with existing dropoffs within 666 meters of the new order's dropoff will be considered for consolidation: 500 * (1 + 5 / 15) == 666

If the distance is 20 kilometers, riders with existing dropoffs within 1100 meters of the new order's dropoff will be considered: 500 * (1 + 20 / 15) == 1100

Riders can be forcibly assigned, ignoring previous rule restrictions, as long as they are within meter_search_radius_whitelist distance. After this assignment, TSP (Traveling Salesman Problem) optimization will be performed.

Under normal circumstances, the number of tasks assigned to any rider cannot exceed nbr_jobs_per_slot.