OVERVIEW:

A customer in the manufacturing industry had started out manually scheduling their process orders, but after years of growth, found the process to be overwhelming. They faced key challenges with:

  • Complexity: With over a thousand unique products and 10 lines, each with their own limitations on which products to produce, scheduling required extensive knowledge of the system.
  • Visibility: Process orders were printed from their ERP system by the sales group and handed to schedulers, who would then physically place them in the proper cubby according to where they wanted it to run. This process did not make it easy to comprehensively see or assess the full schedule.
  • Knowledge: Due to the complexity of the system and lack of a concise handbook for scheduling, the customer experienced issues with internal transfer of knowledge for nuanced operations.
  • Punctuality: Getting orders to the customer on time is key for the business, but challenges in scheduling sometimes led to issues in meeting deadlines.

SOLUTION:

Our process began by learning what made the schedulers tick. What drives their decisions? How do they communicate with production teams? How often do they need to change the schedule?

Together with the customer we worked to develop a set of rules that define how the Smart Scheduler thinks about the scheduling problems put in front of it. This allowed us to create a set of levers the customer could pull on themselves to adjust the decisions the scheduler makes. These adjustable levers included concepts like:

  • Runtimes: Production rates for each product on each viable line, used to determine total runtime for a given quantity in the schedule.
  • Line Preference: Preferential scoring for choosing certain lines for a given product. Also prevents certain lines from being considered for the product.
  • Comparative Weighting: Weighting parameters associated with each internal rule to influence complex decisions on how important a rule is to adhere to when it can’t adhere to all of them. For example, is it better to run a process order slightly later on the preferred production line or run it earlier on its least preferred production line with a forced changeover.

A scheduling system is not just in its brains, however – it also needs to successfully integrate with the rest of the facility to be of use. This customer had a ERP system that handled their process orders, so we integrated with it to pull the new process orders into the system automatically and/or at the push of a button.

Production integration is also key in two respects:

  • Production teams need to know what they’re scheduled to do, and
  • The schedule needs to know about real actions on the floor and respond accordingly.

For the first, we created an assignment feature in the schedule. Assigning work orders locks them in place in the schedule (subsequent rescheduling won’t remove it) and makes it visible to production teams on custom-built production pages that give them additional information about how to execute the work order. This let schedulers have a “draft” schedule and an active schedule coexist and makes sure the production team has a consistent schedule available to them.

For the second, we tied back feedback from real production events so that the schedule could update based on new estimated run times and inform the scheduler. This let subsequent rescheduling consider the new timing in its decisions.

Finally, we made sure the application made it easy to get information each group needed. Visualizations clearly identified different types of products, punctuality in the schedule, status and more so that users could review the automatically generated schedule with ease. While we’d love for the schedule to be followed every time, we also know that sometimes things don’t go to plan, so manual override tools were added to let schedulers manually move work orders around and production leads run work orders out of order or on different production lines.

RESULTS:

Once the Smart Scheduler was implemented onsite, the results were:

  • Total active production hours shortened by 34% due to fewer changeovers and improved production line selection.
  • Production time span shortened by 5 days with reduced production hours plus improved line balancing.
  • Improved order fulfillment time by scheduling items in a better order

They haven’t gone back.

* Results were generated by comparing a two-week schedule manually generated to one automatically generated as part of the validation process prior to shifting to regular use of the smart scheduler.

 

Let's improve your industrial performance