Assembly Verification

Use Cases

Ensuring the correct Assembly

In the manufacturing of finished products, it is very important that the parts are properly assembled.FotoNow checks that the correct parts have been attached, that clearances have been maintained, and that no parts are missing.

Identification of multiple Defects in a Car Door Assembly Line

Industry: Automotive

Problem: The manufacturer was looking for a solution to the problem of error detection in a high-volume manufacturing process of car doors. The manufacturer’s goal was to recognize 20 types of flaws with one solution. The task was made more difficult by the movement of the assembly line during the visual inspection process. Another challenge was an inspection in conditions of high-frequency vibrations.

Solution: The Company installed FotoNow on three smartphones placed on the front, top and bottom sides of the door. The model was trained to identify multiple objects within one picture. The application was configured to be able to recognize all the types of defects defined by the customer.

Results: Device setup took less than one day, and the customer was able to identify all 20 defects with more than 99% accuracy in just one month. In the same timeframe, more expensive competitors were only able to achieve 90% accuracy. FotoNow is now reporting five defects per day in real-time, saving the customer €450,000/year in reduced rework expenses and increased productivity with just one quality checkpoint.

Quality Inspection in a multi-model Production Line

Industry: Automotive

Problem: A car manufacturer has a high-volume body part assembly line. The parts of different car models are assembled at the same time. The parts of different models may look similar, but they have different configuration requirements. The manufacturer could not automate the inspection of the assembly process, because there was no solution that could identify the different car models and therefore correctly analyze the configuration of each part.

Solution: Our algorithm was trained to precisely identify all the car models based on images of certain parts. When those parts passed through the assembly line, FotoNow analyzed their set based on the requirements for each model. If the errors in the assembly were detected, notifications were sent to 60 relevant employees in real-time across departments and geographies.

Results: FotoNow inspects 200 part positions within one image taken with a smartphone camera. Due to its accuracy and quick operation, FotoNow saves the customer more than 250.000 € of rework per year for each inspection station.

Read about more Use Cases