Early SMT Defect Detection with Vanti

Increase throughput with early-detection predictive analytics in manufacturing

Predictive analytics in electronics manufacturing

The cost of SMT defects increases significantly as PCBs progress through the various manufacturing processes. In some cases, when an issue is found far down the line as part of regular quality management processes, it is no longer reversible, resulting in the unit being sent to the scrap heap. In other cases, fixing an item at a late stage may be possible but too expensive. Beyond the direct costs of the wasted material, finding errors late in the production process affects line yield. A solution that enables issues to be found early on in the production line can improve the first-pass yield calculation, reduce costs and improve yield, resulting in high throughput and high ROI.

With the Valor-Vanti Predictive Analytics Solution, we shift left SMT defect prediction, predicting faulty units earlier in the process, and thus reduce the cost of error.

The solution works by analyzing historical test measurements to create an early detection AI model. By training the AI model on the first test iterations, the solution can predict in real-time which units are most likely to fail at a later stage of the manufacturing process. Using predictive analytics in manufacturing, the units can be removed from the production line early on and sent for rework. As a result, a lower percentage of products will be faulty in the latter stages of production, resulting in higher throughput and productivity with improved quality and less waste.

The Valor-Vanti Predictive Analytics Solution is available as an annual subscription plan per product line (SKU), according to data volume and required latency. For high-mix, high-volume sites, usage-based pricing may be applied. Contact your local representative for pricing information.

Early SMT Defect Detection with Vanti

The cost of SMT defects increases significantly as PCBs progress through the various manufacturing processes. In some cases, when an issue is found far down the line as part of regular quality management processes, it is no longer reversible, resulting in the unit being sent to the scrap heap. In other cases, fixing an item at a late stage may be possible but too expensive. Beyond the direct costs of the wasted material, finding errors late in the production process affects line yield. A solution that enables issues to be found early on in the production line can improve the first-pass yield calculation, reduce costs and improve yield, resulting in high throughput and high ROI.

With the Valor-Vanti Predictive Analytics Solution, we shift left SMT defect prediction, predicting faulty units earlier in the process, and thus reduce the cost of error.

The solution works by analyzing historical test measurements to create an early detection AI model. By training the AI model on the first test iterations, the solution can predict in real-time which units are most likely to fail at a later stage of the manufacturing process. Using predictive analytics in manufacturing, the units can be removed from the production line early on and sent for rework. As a result, a lower percentage of products will be faulty in the latter stages of production, resulting in higher throughput and productivity with improved quality and less waste.

The Valor-Vanti Predictive Analytics Solution is available as an annual subscription plan per product line (SKU), according to data volume and required latency. For high-mix, high-volume sites, usage-based pricing may be applied. Contact your local representative for pricing information.

Valor Predictive Analytics

Graphic of the Valor Micro-Solutions cloud

Valor Predictive Analytics Solutions focus on solving small problems in the production line, one step at a time. Utilizing machine learning and artificial intelligence improves yield, quality, and OEE, to solve efficiency challenges.

阅读更多信息

对 [[global-preference-center-interest-placeholder]] 感兴趣?

订阅 [[global-preference-center-interest-placeholder]] 更新信息

哦!

抱歉,我们无法捕获您提交的信息。请重新尝试。

感谢注册!首次订阅?确保选中电子邮件以确认订阅。

字段为必填 必须填写有效的电子邮件地址
字段为必填 值无效或太长