An automatic optical inspection system checks wafer sheets in the moving process for breakage and baked-in pieces. A narrow window between two cord conveyors suffices to ensure 100 % inline quality control.
Sweet wafers are a popular treat – whether pure or filled, as wafer bread or in chocolate bars, ice cream sandwiches or other confectionery. Combined with cream fillings and chocolate layers, their appeal is in their crisp, airy texture. As the adage goes, you eat with your eyes first. Therefore, manufacturers check the baked wafers for uniform appearance and any breakage before cutting, processing and packaging. Common defects are broken edges and baked-in crumbs or broken bits from previous batches.
The former ones show up in transmitted light as bright pixels, the latter ones as dark pixels. This inspection task is therefore well-suited for automation, especially, since machine vision systems are significantly faster and more accurate than human-eye inspection. Furthermore, they can check 100 % of all manufactured wafers without the need to reduce the line speed. Thus, the inspection quality in automated baking lines can be significantly increased by introducing image processing technology.
The Berlin-based machine vision integrator Bi-Ber manufactures automatic optical inspection (AOI) systems for various tasks and quality criteria in the confectionery industry. One new system, designed for a wafer bread manufacturer, checks 470 mm x 290 mm wafer sheets as they leave the oven – one every two seconds. For shade-free images of the wafers, two cord conveyors were set up with a 100 mm wide gap in-between. The inspection system takes four overlapping images of each wafer through this viewing window, evaluates them in real time and, if necessary, initiates the discharge of wafers via pneumatic cylinders.
The pictures are taken in the moving process. The first image of every wafer is triggered via a light barrier, the rest are taken in fixed time intervals. The inspection system distinguishes between pass, fail, and rework. Wafers from the rework stack are at this point in time still being checked manually in order to see if they are at least partially suitable for cutting. This step is suitable for further automation in future by transmitting the error coordinates. The AOI system, moreover, documents the brightness of the wafers and thus the degree of browning. This data can be used as an additional quality criterion and for further automation of the entire line, for example, to adjust furnace settings more dynamically.
The system design not only allows for easy and flexible integration in production lines but also features very easy usability – once switched on, the inspection application runs autonomously. The vision system automatically finds the areas of interest and saves a result image for each wafer, composed of the four partial shots. Different user levels allow for adjusting the tolerances for scrap and rework. Pass/fail statistics are continuously being displayed and archived. The stainless-steel system consists of a GigE camera above the conveyor line, an LED assembly that illuminates the wafers from below, and an operating terminal.
