中文 | English

This is a demo system to dispatch services in London area:

• Address：https://planner.kandbox.com
• user_email: demo
• password: Demo1234

👍 Timeline Chart with travel time

👉 Reinforcement Learning Environment per standard of Gym

👉 Heurisitc Agent for the Environment

It is recommended to use docker version. For more details about installation, click get_started_locally.

1. Prepare linux or macos with docker, docker-compose. This is how to install docker-compose

2. Run this commond in {KPlanner_HOME}/src folder:

   cd src
   docker-compose -f prod.yml build
   docker-compose -f prod.yml up

3. If there is no error, you should be able to create a super user with:

   cd src
   docker-compose -f prod.yml  run backend python manage.py createsuperuser

4. Open a browser and visit http://localhost:8000. You can create the rest of users with the super user credential.

5. (optional) To populate sample data, run :

   docker-compose -f prod.yml  run backend python kandbox_planner/fsm_adapter/toy_generator/london_service_generator.py 	<your_access_token>

Reinforcement Learning, Field Service Scheduling, Dispatching, Planning, AI, Optimization

Frontend UI: Django | Django-SimpleUI

Backend Algorithm Platform: Gym | RLLib for RL | Ortools for Optimization

Everything is in python but JS in browser.

• Improve London area dispatching quality
• Create usage documents