Understanding Forecast Changes and Lags
Understanding the root cause of forecast volatility is important in Motorola Solutions’ complex configure-to-order business, which involves thousands of different components.
Motorola Solutions recently worked with Northwestern’s Industrial Engineering (IEMS) Client Project Challenge students to help them understand the root cause of forecast changes.
The project focused on one of Motorola Solutions’ configure-to-order sites. Motorola Solutions’ customers order products to suit their needs, which means that they must forecast the demand for each of the thousands of potential components. The components can all have different lead times.
Chad Werkema, Supply Chain at Motorola Solutions said, “We wanted to give the students a problem that reflected the complexity of the real world.”
As with any supply chain, understanding the root cause of forecast volatility can be difficult because of the three dimensions:
1. The forecast is updated weekly.
2. The graph below visualizes the problem. For a given component on a given day, there is a forecast (like the orange line below).
3. Then, next week, a new forecast will be run (the gray line, then the yellow line).
The project’s deliverables included data engineering to pull the data together for analysis, building a dashboard to visualize the findings and analytics to help pinpoint the root causes of added volatility.
Ken Milleville, Supply Chain at Motorola Solutions said, “I was impressed with the analytical capability of Sophia, Ali, Grant and Rhea. They impressed me with their ability to understand the data and apply analytics. It made me realize that I’ll need to give the student team a bigger challenge next year!”
Congratulations to all involved, especially the student team (Sophia Dorn, Ali E., Grant Galatas and Rhea Shah) and their faculty advisor, Professor Ermin Wei.
