Large manufacturers depending on the industry, produce hundreds to thousands of units per day. In the automotive industry, for example, approximately 1,000 units leave a single production plant every day. And how many of them turn back? Roughly 8% of produced cars are recalled because of poor quality [1]. To the most product groups in manufacturing being recalled belong, except cars, construction materials, electronics and medical devices. According to Allianz Group, each major recall costs a manufacturer more than 10 million dollars [2]. In addition to enormous costs directly associated with the recall, there are also penalties and fines. Another serious consequence of a product recall is the ruination of the brand image. Product recalls belong to the top 5 causes that have an impact on reputational issues for brands [3].

The root cause of all the recalls is quality management issues, which includes many factors such as the quality of machinery, training of skilled workers, production planning and quality control. According to the study on the cost of poor quality, about half of all defects in production can be detected by visual quality control [4]. Accordingly, inspection of visual defects can significantly minimise the likelihood of a product recall. Faults that can be examined by visual inspection include scratches, dents, cracks, misplaced or missing parts, color defects and deformations. For many products, even tiny flaws are critical to functionality and safety.

Manufacturers, of course, execute visual quality assurance. But why do so many defects still occur? In most companies, even large ones, quality control is performed manually, because they produce personalized products or change their product line frequently. Thus, they cannot use existing solutions for machine defect detection. Current automated visual quality control vendors require months of training each time a product change is made. Given this, the company’s employees must look at each finished part and evaluate whether it has defects. Highly skilled employees are under-challenged with such tasks and also fail to deliver accurate results because the human eye is prone to error. However, even those companies that have stable and non-complex product lines and use existing visual quality inspection solutions, have quality issues. Issues occur mainly because the lighting conditions, line movement and the vibrations in a manufacturing plant are unstable. Therefore, the cameras cannot capture sharp images, which consequently leads to incorrect defect analysis.

To assure producers that defective parts will not pass through the visual quality inspection station, FotoNow has developed software that is flexible and accurate. First, we have developed a technology thanks to which the program can be trained within hours, not months. The setup requires only 5 images and is simple so that the solution is implemented within a day. This makes it possible to replace the quality inspection, which is currently performed manually, with the automatic one and get much more accurate results. FotoNow is flexible not only in terms of changes to the product line but also in terms of functionality in the real manufacturing conditions. Due to its enhanced image capturing model, FotoNow is able to perform precise quality inspection in conditions of changing lighting, high-frequency vibrations and objects in motion. Moreover, it does not have to be fixed to a certain position, so you can relocate the camera, and our solution will detect the faults in the entire field of vision with the same precision. Furthermore, our system allows image inspection on a pixel level which helps inspect errors with more accuracy and differentiate defects from false-positives (i.e. dust, fingerprints, reflection, etc.). 

FotoNow brings demonstrable improvement in quality control to producers. If you want to read more about our use cases, follow the link: .


[1] Handelsblatt. (20. August, 2019). Anzahl der zurückgerufenen Fahrzeuge nach ausgewählten Hersteller im 1. Halbjahr 2019 in Deutschland.

[2] Allianz SE, Was kostet ein Produktrückruf? ( 

[3] Harris Poll, 2016 Reputational Quotient. ( 

[4] S. N. Teli et al, 2013: Cost of Poor Quality Analysis for Automobile Industry.



by | Jan 21, 2021