ReinforcementScheduler

A mobile application that helps Chevron’s management team track and assign work orders to available and eligible workers through reinforcement learning.

Comment

Functions

The application has two major functions - tracking existing tasks(categorized into Unassigned and In-progress), adding new tasks or workers to the profile, updating the workers' and tasks' information according to worker capabilities and conditions and tasks' attributes, and displaying logging information based on the operations in the app.

How We Built It

The mobile application we created has a strong and advanced back-end based on deep reinforcement, dynamically modeling the problem based on part of Markov decision process and optimally adjusting the schedules in real time. We built the frontend using Swift.

Challenges

We encountered the problem of not being able to push our new commits to GitHub because of failures to merge files several times during our hack time. We figured out that it was caused by changes to the project that involved the creation of several new ViewControllers, made by different people simultaneously. Another problem we ran into involves the integration between front and back end.

Built With

Share this project:

Updates