By: Bala Deshpande, Conference Co-Chair, Predictive Analytics World for Manufacturing 2015 

In anticipation of his upcoming Predictive Analytics World for Manufacturing conference Jeffrey_Thompsonpresentation, Manufacturing Analytics at Scale: Data Mining and Machine Learning inside Bosch, we interviewed Jeffrey Thompson, Senior Data Scientist at Robert Bosch, LLC. View the Q-and-A below to see how Jeffrey Thompson has incorporated predictive analytics into manufacturing at Robert Bosch, LLC. Also, glimpse what’s in store at the PAW Manufacturing conference.

Q: In your work with predictive analytics, what behavior do your models predict?

A: At Bosch we work on a wide variety of different use-cases.  The target applications of our predictive models include manufacturing, supply chain and logistics, engineering, and Internet of Things and Services.    

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A: Predictive models are used across products, processes, and operations at Bosch. An example use of our predictive models is in providing important insight into the root causes of failures on manufacturing lines.  These insights often lead directly to the resolution of a problem and are an important part of our continuous improvement efforts.

Q: Can you describe a successful result, such as the predictive lift (or accuracy) of your model or the ROI of an analytics initiative?

A: In one case, we used a predictive model to narrow down the root cause of a particularly expensive, internal defect in one of our automotive manufacturing lines.  The relative improvement in this case was 85%.   

Q: What surprising discovery have you unearthed in your data?

A: Data often tells you many things that you are not looking for and hence offers many surprises in all our projects. For example, many data sources or measurements never change over the course of a product’s or a process’ lifetime. It might be expensive to measure them, yet they continue to be measured based on an initial and incomplete design specification.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World for Manufacturing.

A: In many of our applications, high accuracy, and especially, low false alarm rates are critical. In such scenarios, even with large amounts of data, we find that better features often beat better algorithms, and subject matter experts are the key to getting those features. 


Don't miss Jeffrey Thompson’s conference presentation, Manufacturing Analytics at Scale: Data Mining and Machine Learning inside Bosch, at PAW Manufacturing, on Tuesday, June 9, 2015, from 2:40-3:25 pm. Click here to register for attendance. 

By: Bala Deshpande, Founder, Simafore and Conference Co-Chair of Predictive Analytics World for Manufacturing.