Machine learning is a discipline of artificial intelligence that enables computers to learn from data and improve performance on specific tasks without being explicitly programmed.

Through data analysis and pattern identification, machine learning algorithms provide predictions, detect patterns, and in some cases make autonomous operational decisions. The application of machine learning is revolutionizing numerous fields, including medicine, finance, industrial automation, electronics and many others.

A striking example of this transformation is represented by a company in the Northwest of Italy, active in the industrial electronics sector, which, thanks to Widecons SRL ®, has been able to successfully face challenges related to quality control thanks to innovative solutions based on artificial intelligence.

Initial problem:

The company was facing a critical issue related to automated visual inspection of printed circuit boards supplied by subcontractors.

With an average failure rate of 3.2% and a high failure rate at the next functional test (9.8%), product efficiency and quality were compromised.

Widecons intervention:

To address this challenge, the company turned to Widecons, which, through a tailored approach, implemented a machine learning algorithm dedicated to improving visual inspection, with the goal of achieving progressive improvements over time.

Implemented solutions:

Widecons adopted a Smart Small Solution to define the problem and identify the most suitable solution, followed by a Customized Solution to develop the machine learning algorithm.

Results:

Just four months after implementing the new algorithm, the company has seen a significant improvement in results.

The effectiveness of visual inspection increased, reaching an average PCB failure rate that rose to 6.3 percent, with a significant reduction to 3.2 percent in the subsequent functional test. Over the course of one year, these improvements continued, with the failure rate at 7 percent in visual inspection and an indicative 1.5 percent in functional testing.

In addition, with the increased clarity of data obtained from visual inspection, the company was able to work with the supplier to improve PCB quality directly at the source.

This led to the subsequent decentralization of visual inspection to suppliers and the elimination of in-house inspection, reducing lead times, increasing productivity and the overall quality of PCBs and the finished product.