CASE STUDY: Manufatturer of beverage

Reduce the cost of full-surface beverage cap inspection system to 1/5 by using AI




Originally, inspections were made with inspection equipment from other makers that used rule-based algorithms. As there were restrictions on alignment, large-scale equipment was needed to perform inspections. Further, in order to see all of a cap’s outside and underside, these devices came equipped with 14 cameras. For damage inspections of a cap’s interior, such equipment used X-rays.

Issues associated with cap inspection equipment

There were three main problems with cap inspection equipment

  • Alignment restrictions were quite rigid due to inspections being made with rule-based algorithms, and large-scale equipment was needed.
  • With 14 cameras installed, maintenance was complicated, resulting in prolonged suspensions of equipment use.
  • Costly X-ray methods had to be used to make damage inspections of caps’ interiors.

Realize perimeter imaging with special lenses and automatic alignments using AI

When inspecting a cylindrical workpiece such as a milk cap, it used to be necessary to take images of the inside and outside of the workpiece from multiple directions with multiple cameras. However, by using a special lens, it is now possible to take images of the entire circumference with a single camera. This simplifies the layout of the vision system and speeds up the inspection process, as it eliminates the need to take images multiple times and rotate the workpiece. Furthermore, by using AI, automatic positioning, which was difficult with conventional rule-based algorithms, has been achieved, and the system has been downsized and made less costly.

Reduce inspection equipment costs to 1/5th or less of what they were

Through these solutions, inspection equipment that used to cost 100 million yen or more per unit was reduced to 20 million yen or less.