This article was written because of great interest among QA staff and Route4Me users about compability between the results from defferent applications:
- web UI. Results of route optimization on the official website: https://www.route4me.com;
- curl (cmd). Results of route optimization using curl (cmd);
- Excel add-in. Results of route optimizatin using Excel add-in.
For the aplications we prepared input data with similar content:
Parameters:
"parameters": {
"algorithm_type": 1,
"store_route": false,
"route_name": "Single Driver Multiple TimeWindows 12 Stops",
"route_time": 25200,
"optimize": "Time",
"distance_unit": "mi",
"device_type": "web",
"vehicle_max_distance_mi": 1000,
"vehicle_capacity": 1,
"rt": true,
"route_max_duration": 86400,
"travel_mode": "Driving"
}
Addresses:
"addresses": [
{
"address": "1201 Holcombe Blvd, Houston, TX 77030, USA",
"alias": "Start",
"is_depot": true,
"lat": 29.7063457,
"lng": -95.3999784,
"time_window_start": null,
"time_window_end": null,
"time_window_start_2": null,
"time_window_end_2": null,
"time": null
},
{
"address": "1220 Holcombe Blvd, Houston, TX, 77030",
"alias": "Betty Thornburg",
"lat": 29.7055795,
"lng": -95.400029,
"time_window_start": 28800,
"time_window_end": 43200,
"time": 300
},
{
"address": "6651 Main Street, Houston, TX, 77030",
"alias": "Corey Barker",
"lat": 29.708566,
"lng": -95.4024123,
"time_window_start": 32400,
"time_window_end": 39600,
"time": 100
},
{
"address": "6100 Main St, Houston, TX, 77005",
"alias": "Evon Caldwell",
"lat": 29.7184075,
"lng": -95.3980766,
"time_window_start": 28800,
"time_window_end": 54000,
"time": 300
},
{
"address": "1520 Louisiana St, Houston, TX 77002, USA",
"alias": "Jody McDougall",
"lat": 29.7543582,
"lng": -95.3708402,
"time_window_start": 32400,
"time_window_end": 64800,
"time": 300
},
{
"address": "6633 Travis St, Houston, TX, 77030-1308",
"alias": "Nancy Chiu",
"lat": 29.709903,
"lng": -95.403276,
"time_window_start": 30600,
"time_window_end": 57600,
"time": 300
},
{
"address": "4001 Hazard St, Houston, TX 77098, USA",
"alias": "Patrick Allison",
"lat": 29.7366671,
"lng": -95.4068916,
"time_window_start": 28800,
"time_window_end": 54000,
"time": 300
},
{
"address": "6720 Bertner Ave, Houston, TX, 77030",
"alias": "Sandra Sabatini",
"lat": 29.707876,
"lng": -95.400405,
"time_window_start": 23400,
"time_window_end": 64800,
"time": 300
},
{
"address": "1515 Holcombe Blvd, Houston, TX, 77030-4000",
"alias": "Tarsha Wilson",
"lat": 29.7071962,
"lng": -95.3966905,
"time_window_start": 25200,
"time_window_end": 61200,
"time": 300
},
{
"address": "1 Baylor Plaza, Houston, TX, 77030",
"alias": "Tricia Magana",
"lat": 29.7105286,
"lng": -95.3962414,
"time_window_start": 21600,
"time_window_end": 46800,
"time": 300
},
{
"address": "1220 Blalock Rd, Ste 150, Houston, TX, 77055",
"alias": "William Espinoza",
"lat": 29.7903957,
"lng": -95.5213012,
"time_window_start": 27000,
"time_window_end": 64800,
"time": 300
}
]
Note: These addresses are not allowed for demo account with api_key=11111111111111111111111111111111.
Response JSON file - web UI Single Driver Multiple TW
The route one the map.
Response JSON file - curl (cmd) Single Driver Multiple TW
The route one the map.
Response JSON file - Excel add-in Single Driver Multiple TW
The route one the map.
The results from curl (cmd) and Excel add-in are equal. Ther result from web UI is same but movement on the route is opposite.
For this optimization type was changed only the parameter algorithm_type:
algorithm_type=3
The addresses are same as in the previous optimization type.
In case of web UI we got 1 route.
Response JSON file - web UI Multiple Driver Multiple TW
The route one the map.
Was got 10 routes - from depot to other stops one by one.
Response JSON file - curl (cmd) Multiple Driver Multiple TW, Vehicle Capacity = 1
Was got 10 routes - from depot to other stops one by one.
Response JSON file - Excel add-in Multiple Driver Multiple TW, Vehicle Capacity = 1
The results from the curl (cmd) and Excel add-in are equal again. The web UI result is different - it seems low value of the parameter vehicle_capacity didn't affected on the result - altough, in the response file we see that
vehicle_capacity=1
The parameters and addresses are same except the parameter vehicle_capacity
vehicle_capacity=30
Results from web UI is same as vehicle_capacity=1
Response JSON file - web UI Single Driver Multiple TW, Vehicle Capacity = 30
The route one the map.
Response JSON file - curl (cmd) Single Driver Multiple TW, Vehicle Capacity = 30
The route one the map.
Response JSON file - Excel add-in Single Driver Multiple TW, Vehicle Capacity = 30
The route one the map.
The results from curl (cmd) and Excel add-in are equal. Ther route shape from web UI is same, but the sequence of the addresses is different.
The name for the value 3 of parameter algorithm is CVRP_TW_SD - Capacitated Vehicle Routing Problem, Time Windows, Single Depot. Of course, this algorithm assumes that we should pay attention to the parameter vehicle_capacity and set it to right value. If we'll set it to 1, algorithm will be forced to generate almost as many routes as many stops are in the optimization.
Please, pay attention to this table, which shows corespodence between API parameters and web UI field names.
| web UI field | API parameter |
|---|---|
| Maximum Pieces per Route | vehicle_capacity |
| Maximum Weight per Vehicle | vehicle_max_cargo_weight |
| Maximum Volume per Route | vehicle_max_cargo_volume |
| Maximum Distance per Vehicle or Route | vehicle_max_distance_mi |
| Maximum Number of Routes/Vehicles | parts |
| Maximum Stops per Route | max_tour_size |