Quicksilver — Courier Routing Optimization Service

A lightweight Go service for solving the Vehicle Routing Problem (VRP) — optimal assignment of delivery tasks to couriers, considering geolocation, time windows, and capacity constraints.

Available Algorithms

• Greedy Assignment Algorithm

Features

• Accepts a list of couriers and tasks via HTTP API.
• Assigns tasks to couriers using a greedy algorithm.
• Takes into account:
• Distance between points (Haversine formula).
• Time windows (pickup and delivery).
• Courier capacity (volume and weight).
• Task durations (pickup and drop-off times).

API

POST /vpr/greedy/

requset:

{
 "couriers": [
   {
     "guid": "c1",
     "start_point": { "lat": 55.75, "lon": 37.62 },
     "capacity": { "volume": 10, "weight": 20 },
     "pickup_duration": 120,
     "drop_duration": 60
   }
 ],
 "tasks": [
   {
     "guid": "t1",
     "sender_point": { "lat": 55.76, "lon": 37.60 },
     "recipient_point": { "lat": 55.78, "lon": 37.65 },
     "capacity": { "volume": 2, "weight": 3 },
     "assembly": {
       "from": "2025-04-05T10:00:00Z",
       "to": "2025-04-05T12:00:00Z"
     },
     "slot": {
       "from": "2025-04-05T11:00:00Z",
       "to": "2025-04-05T14:00:00Z"
     }
   }
 ]
}

response:

{
 "routes": [
   {
     "courier_guid": "c1",
     "route": [
       "t1"
     ]
   }
 ],
 "unassigned": null
}

TODO

• More advanced distribution algorithms
• More options for fine-tuning