Challenge: A major fast food manufacturer requires their bakery to control the variation of the bun height on a new product. Excessive variation will cause the stores to vary the toaster settings which will result in inconsistent browning of the bun and long cycle times in the stores. The Cpk of the bun height variation is averaging .60 and should be 1.33.

Outcome: The dough weight and the dough temperature are the most significant factors that affect the amount of variation of the bun height after baking. Excessive variation from the inputs to the process and packaging/equipment issues are the main causes.

Recommendation: Conduct projects to control the inputs to the process to a Cpk of 1.33. Focus on the dough weight and dough temperature first. Conduct a risk analysis of the product line utilizing Failure Modes and Effects Analysis to minimize downtime and special causes of variation (equipment failure, wearout,etc).

The Process: The Define, Measure, Analyze, Improve and Control (DMAIC) project model was followed. First, a project charter with mission, business case and timelines was created. Top management approved the charter and a team was formed.

• Define Phase: The existing conditions were analyzed, documented and understood. A baseline run chart of the Cpk, means, standard deviation, range and percent defective were constructed over an 8 week time frame.
• Measure Phase: The Design of Experiment (DOE) tool was selected as the best method to evaluate the cause of the variation due to the number of factors and interaction effects of the different steps in the process. An experiment was initiated to analyze the dough weight, dough temperature, floor time and oven temperature at different levels and data was collected. The output response was the height of the bun after baking.
• Analyze Phase: After all the runs were completed (16) and data collected, Minitab Software was used to analyze the results. The software predicted that by optimizing the various factor settings, an 80% improvement would be realized.
• Improve Phase: The line was set to the optimum settings as determined by the software and a confirmation run was conducted and the predicted improvement was verified.
• Control Phase: The inputs will be controlled to minimize their variation and the special causes will be eliminated. Statistical Process Control Charts will be used to monitor this variation.

The project determined the significant factors and causes of the variation to the bun height dimension. Further projects are required to determine the corrective actions for these causes.