HRC painting re/scheduler
General Description
The HRC scheduler is designed to help incorporate collaborative robots into conventional furniture painteing plant. This advanced tool offers a streamlined and interactive interface for calculating, visualizing, and adjusting both weekly and daily schedules, integrating cobot training as an essential component of the painting process.
| Resource | Link |
|---|---|
| Source code | Scheduling module API |
| Demo Video | Link to video |
Contact
The following table includes contact information of the main developers in charge of the component:
| Name | Organisation | |
|---|---|---|
| UPV | mterol@cigip.upv.es |  |
| UPV | lmoya@cigip.upv.es | |
| ITI | mgerar@iti.es |  |
Integrated and Open Source Components
This section outlines both pre-existing technologies and open-source components that have been integrated into the HRC painting re/scheduler. It provides details on their sources, purposes, modifications, and compliance with licensing terms.
Overview
The HRC painting re/scheduler is a scheduling application that is integrated with the Collaboration Ambient Data Model database. It provides a seamless task management for painting production, and manage scheduling information easily. The application retrieves the required data from the database, which is used by the scheduler to provide a scheduling proposal for the required painting tasks. Once the schedule is approved, the application stores the scheduling for each task following the result and updates the information in the database. It also manages cobot training sessions when new products require painting programs.
Pre-existing Components
Collaboration Ambient Data Model
Source
The Collaboration Ambient Data Model is available in the AI-PRISM repository.
Description
The data model database serves as the central repository for storing and managing scheduling-related data. It includes records of human and cobot availability, tasks, training requirements and execution data. This structured data allows the scheduler to provide an optimized scheduling proposal for painting tasks.
Modifications
The data model has been adapted to include new fields and tables to support cobots in the scheduling of painting tasks.
Purpose in AI-PRISM
The data model database enables efficient and automated scheduling by providing structured access to processes and resources, ensuring an accurate scheduling based on the company resources and constraints. It also keeps track of cobot training sessions to ensure that the cobots are always up-to-date with the latest painting programs.
License
MIT License
FACOP (Framework for Applied Computational Optimization Problems)
FACOP is a set of libraries developed by ITI to accelerate the process of creating, testing, and deploying optimization algorithms. These libraries provide the necessary methods and components for constructing algorithms by combining existing elements. Each component is specialized in a specific process, such as data reading, solution generation, solution evaluation, solution comparison, objective measurement, among others. FACOP enables the creation of new components while reusing previously developed ones, allowing the construction of new algorithms by integrating different components capable of deriving solutions for a given optimization problem. In this solution FACOP was used to create and compose the scheduling algoritm used in AW pilot.
Modifications
Necessary components to calculate different scenarios schedules where developed, like several components needed to load external data, an three evaluation component to measure the proposed solutions values, two objective components. Some modifications where made to a genetic algorithm. We created and API that published some methods to send data and receive results in form of scheduling tasks. This API can compose the necessary algorithm for weekly and daily scheduling from a given configuration using FACOP components.
Purpose in AI-PRISM
FACOP is used in this project to create and execute algorithms for weekly and daily schedule, and also an algorithm for daily re-scheduling. All these schedules will incorportate the training of the cobot arm as part of the plant scheduling.
License
Prorietary
How to install
This component does not requiere further configuration to be used or installed.
How to use
The HRC painting re/scheduler is accessed through the browser interface. The user can log in with their credentials and access the scheduling application. The main functionalities of the application include both weekly and daily scheduling, which include both production tasks and cobot training sessions. To get the best scheduling proposal, the user needs to input the date, as this process is done weekly on mondays based on the information available by last Friday and new orders. The application will provide a scheduling proposal based on all orders that should be completed in the week, so the user can approve or reject, making the algorithm propose another schedule. Once one schedule is approved and stored in the database, the system will update all information, and the user can access the daily scheduling to check the tasks for the day. The information will be displayed through a visual representation of the tasks in a gantt component and also through a list of tasks with detailed information in a table.
Scheduling Tasks
The scheduling process is an automatic process, so when a new schedule is created the user only has to provide the date so the system get which week is being scheduled. Once it is introduced, all information required is retrieved from the database: available workers and cobots, task priorities, capabilities, and training requirements. After getting a scheduling proposal, the result will appear in the dashboard, where the user can approve or reject it. If rejected, the system will propose another schedule.
Scheduling Cobot Training
A training session is required if a cobot does not have a painting program for a new product.