Case study on optical systems

Automated testing system for ophthalmic lenses

Uvex, Plant Fürth

Initial situation

OLIGO surface controls was commissioned to develop an automated testing system for ophthalmic lenses. In advance, Uvex provided us with several defect samples and good parts; the defect classes were defined based on a defect catalog.

Approach / Challenge

After careful testing in several different hardware setups, we chose a partner for the hardware and GUI to complete the system. The system includes a 5MP high-end camera, a specially modified grazing light, and a controller (industrial PC).

Solution – Test process procedure

The robot brings the glasses/lenses to be tested to the light; due to the curvature, each lens must be tested separately. Within a few milliseconds, several stripe patterns appear. The distance between the camera, which is mounted on the ball head, and the test object is approximately 270 mm, while the distance from the test object to the strip light or the transmitted light unit is approximately 50 mm. The duration of the test process is 10-30 milliseconds, depending on the number of defect classes.

The touchscreen control monitor displays all the important information, including photos of both lenses. Defects are marked with a red or green “X” depending on the defect class; a red “X” indicates a defect. The user interface can be customized.

If no defect is detected, or if the defects are within the tolerance range, the robot places the glasses on the conveyor belt.
If a defect is detected, the robot places the glasses in a designated container. Several containers can be set up to separate the glasses according to defect classes. This makes it possible to subsequently recheck the missing parts manually and adjust the fault threshold, thus minimizing waste.

Result

The system offers the customer the following advantages:

  1. Minimizes the defect rate. The system operates at a consistently high level without fluctuations 24/7.
  2. The system detects an accumulation of certain defects and issues a warning message. This allows the causes of defects to be quickly identified and eliminated.
  3. Defect classes and numbers are evaluated; it is possible to locate causes of defects, such as contamination or defective coatings, at any time and eliminate the causes and log the statistics on the frequency, e.g., per shift, of the defect classes that occur.
  4. Regardless of the system load, controls are conducted at 100%.
  5. This significantly reduces working hours while minimizing defect rates, thus improving profitability. The ROI can be calculated in advance.