3D Inspection of Baked Bread

Problem Statement:

A bakery relies on statistical manual inspection, basically sampling multiple loaves from the production line on a regular basis to detect process issues that affect the quality of the product shipped. These inspections would typically include:

• Inspect the coverage of any toppings or coatings applied, such as grains or seeds
• Measure the size of the loaf
• Inspect the shape of the top of the loaf to detect when the center of the loaf “falls” or an excessive fissure forms during baking
• Detect burn marks

With volume increasing and new types of bread being produced, statistical inspection lead to too many loaves escaping to the consumer that did not meet the established quality criteria.

VAIA Solution Description:

VAIA’s inspection solution mounts over a section of dedicated conveyor and automatically inspects 100% of the loaves of bread after they exit the oven and are de-panned. An inspection “recipe” is created for each bread SKU so that all the appropriate inspection parameters and pass/fail threshold values are automatically loaded when that SKU is loaded.

This solution utilizes 2 types of vision inspection technology:

3D Machine Vision Inspection

• Measures the length, width, and height of the loaf
• Analyzes the topography of the surface of the loaf to automatically detect collapsed loaves or other shape variations that exceed the inspection tolerances established for the 3D shape of the “ideal” loaf
• Inspects for proper toppings distribution and coverage
• In the 3D images below, darker grey-levels represent lower planes and lighter grey-levels represent higher planes

2D Color Inspection

• Inspects the bread color, if visible, to determine whether the inspected surface color is within the expected color range, as determined by the bake time and temperature
• Inspects the topping colors to determine whether the expected toppings have been dispensed
• Detects significant dark discolorations in the loaf or toppings surface, compared to the typical     surface colors, indicating a burn mark or other potential defect

Handling and Reject

• The loaves move along an existing conveyor, with no special handling required
• Loaves that do not meet ALL of the quality criteria established are rejected from the line and do not enter the consumer stream

Technologies Integrated:

• 3D machine vision sensor configuration to provide 3D topographical top surface analysis and measurements
• 2D color vision configuration to provide color analysis
• Inspection recipe creation and management architecture for each loaf SKU
• Imaging and electronics enclosures that meets construction requirements for food-grade installation and wash-down capability
• Integrated rejecter to remove non-conforming product
• Inspection data analysis to detect trending

Benefits

Using the 3D Machine Vision inspection system, the client is able to monitor the quality of 100% of the bread loaves in real-time as they move through production on their way to cooling and packaging. By detecting defects as quickly as possible after baking, it is possible to make corrections to the topping distribution and baking processes quickly, rather than creating hundreds of defective loaves, some of which are bound to escape. In the end, the goal is to prevent sub-standard product from making it to the store shelf.