Description

The video below summarizes the case study solved by this Python model, and demonstrates the exceution of the Python model.

This downloadable virtual product contains mathematical model descriptions for two common machine changeover optimizations, and a template for optimal machine changeover sequencing in Python. The Python model is an adjustable template suitable for any setup matrix with any number of product groups being produced in a single machine, either as:

• one time setup sequence (one product setup as the first, and one as the last, with changeovers in between)
• continuous repeated changeover cycle

The exemplary changeover matrix implemented by the Python model is the one below:

For the first product in the product sequence, in the case of a one-time-only production program, the following setup times are considered by the exemplary case study:

Alongside a Python template the downloadable product contains a case study problem description and two mathematical model formulations, for both the one time setup sequence problem and the repeated changeover problem.

Who will benefit from changeover sequencing with Python

This downloadable product suits production and manufacturing managers, operation excellence managers, supply chain and production planning consultants, and production planners that:

• Need to identify optimal setup and changeover sequences for their machines.
• Want to optimize changeover optimization in Python.
• Want to understand how machine setup sequences are modeled.
• Want to learn mathematical programming for production planning.
• Want to learn more about how mathematical programming can be implemented in Python.

Python, used for implementing the underlying mathematical setup-sequencing model, supports a setup matrix with an arbitrary amount of product types. This is a major benefit over e.g. using Excel Solver for single machine changeover sequencing – even if you current product planning process is implemented in Excel.

Brief description of the machine setup sequencing problem

This Excel template implements an example with eight product groups running on a single machine: A, B, C, D, E, F, G, and H. Between each product group changeover times differ. For example, preparing an offset printing machine depends on the previous and next printing job.

The user defines setup times and changeover for each setup pair, “from” (previous job) to “next” (upcoming job) product. Each product is setup once, and the model optimizes (i.e. minimizes) the total setup time for a complete setup cycle. That is, a complete sequence where there is a first setup for the first product (no previous product) and setups between all other products.

Two Python models are provided in this downloadable product, implementing two slightly different changeover sequencing problems:

• One time product program, where each product is only produced once – meaning that one product is setup first (setup time), and one is the last product, with every changeover in between being defined with a changeover time.
• Repeated changeover matrix, i.e. each product is produced repeatedly throughout time, and thus a changeover cycle has to be identified (no “first” and “last” setup, instead repeated changeover cycle);

Relevant KPIs considered and outputted by this model

The following KPI is supported by the model:

• Total setup and changeover time for a complete sequence (every product produced once).

Learn more about production planning and analytics

Optimal setup sequencing using mathematical programming is just one example of how you can use mathematical programming in daily operations. You can find many more application examples in this shop and on the SCDA blog.

Here are some other inspirational articles for you to read:

• Link: Job shop scheduling challenges
• Link: Genetic job shop machine task scheduling
• Link: PUSH production and analytics
• Link: Fractional job scheduling