Factory Insights:
25 Years of Data Expertise Focused in one PCBA-Specific Analytics Platform

Factory analytics goes far beyond simple data visualization and dashboards. Sure, these tools provide value by illustrating a live view of the factory, but a true analytics platform should offer far more than nice graphics. The ability to properly analyze trends, including line or machine performance comparisons, as well as to easily uncover hidden performance issues, with the capability to drill down into problems – this is what your analytics platform should aim for. However, this is far easier said than done.

With a reputation established over the past 25 years as specialists in manufacturing data, the Cogiscan team has applied this extensive expertise to a closely related domain: factory analytics – culminating in the creation of a new product, Factory Insights. As a performance analytics platform, Factory Insights takes data collected from the entire production ecosystem, including from machines, sensors, traceability tools (material and process information), and through an underlying data management engine, transforms that rich data into KPIs with actionable intelligence. With Factory Insights, Cogiscan continues to leverage all that data know-how and experience to help manufacturers run a more efficient, timely and thus less expensive operation.

Some KPIs can be accessed directly from the machine, but many need additional contextualization to be calculated accurately. This is particularly important when trying to get relevant factory and line level KPIs. Most BI platforms do a good job visualizing data, but in an industry with as much intricacies as electronics manufacturing, they lack context, or the ability to sanity check the data appropriately.

As an example, we can think of Google Analytics, the most widely used analytics platform for marketing. What makes Google Analytics far better than their competition is that they have a complete understanding of where the data is coming from and how it should be interpreted… because well, they are more than likely the company who originally created and extracted the data. Without a doubt, their dashboards are great, but what really stands them apart is their ability to make sense of the vast amounts of data they’re getting and presenting clear and understandable analyses.

The same applies to PCBA manufacturing, and there is a deep chasm between data visualization and data analytics. Cogiscan is well positioned to contextualize and enrich production data, because they’ve been connecting, accessing, processing, and sharing shopfloor data in electronics manufacturing since 1999. Solely focused on electronics manufacturing, Factory Insights includes a data enrichment & KPI calculation engine specifically designed to handle data for PCB assembly. Sitting close to the shop floor, Factory Insights differentiates itself from other solutions with its ability to transform structured and unstructured data into both sensical and human readable KPIs. From there, anything is possible: you can either use and modify Factory Insight’s pre-built dashboards, or feed the relevant KPIs to any other 3rd party analytics/BI tool.

As mentioned previously, most standard analytics solutions are unable to capture all the nuances related to PCBA manufacturing, which can render the reports both unreliable and unactionable. In other words, having data is a great starting point, but the important part is what you can do with it. As you know, baseline metrics are impacted by a lot of details (double-sided PCBA, dual lane, recipes, AOI re-runs, etc.) so having standardization is of the utmost importance, especially when you want to perform comparisons or drive change. For example, it’s nice to know that one line performed twice as fast as the other, but it means nothing if those lines were running completely different product types. And this type of insight is only available if your analytics platform understands the underlying data. In fact, any line or site comparison is only possible if your analytics platform recognizes the definition of a line and how all the machines and product types are configured.

Time comparisons suffer from a similar problem. It’s great to know that the placement machine’s OEE is now at 84%, but it’s far more valuable to know that placement machine’s overall trend. Is 84% a fluke? And which direction is it going in, up or down? Another example is when comparing the defect rate between two orders on the same line – the data is irrelevant if the calculations underneath were performed differently. In other words, standardization is key for any comparison, and standardization comes from the ability to contextualize and interpret data in varying situations. We’ve all heard the phrase, “numbers can be manipulated to say whatever you want”. Sure, but we all know that’s not what’s needed. KPIs need to be anchored in the same source of TRUTH. And to establish this appropriately, the vast amount of data you’re collecting from the factory needs to be standardized and contextualized.

The underlying engine that supports Factory Insights has been engineered to handle all the intricacies of PCBA manufacturing, while supporting the option to correlate data with other 3rd party systems like MES/MOM or TTC platform.

Machine-level data analysis tools can provide a lot of value in regard to the specific intricacies of that machine’s performance and capabilities, but if looked at in a narrow perspective, can often miss the overall impact the entire line can have on certain machine-level metrics. For example, if an AOI on Line 2 is showing a low OEE, it’s important to consider that multiple factors could be driving it well beyond the AOI machine itself – it could be induced because of a bottleneck at the placement machine, or even at the reflow oven.

In cases where you have to switch from one analytics tool to another, such as jumping from one machine-level analysis tool to another, it can cause headaches related to the standardization and comparison problems explained above. Factory Insights offers a single platform to collect data from all machine types, correlate and enrich that data – enabling you to go from high level OEE all the way to the placement performance of a specific machine’s nozzle all in one platform.

The multi-platform can also make root cause analyses cumbersome and manually intensive. We all know how resource intensive quality investigations can be and without visibility on common reasons and how to prioritize them, too much time can be wasted toggling between multiple systems and reports. With Factory Insights, quality engineers can query consolidated data easily and create custom reports – not limited by real-time data, historical comparisons and analyses are also available. With a single source of truth in one platform, the time to solve quality complaints and warranty returns can be reduced dramatically.

We all know how hard it is to access to equipment data. The electronics manufacturing industry has a wide range of proprietary machine communication technologies and standards, which translates into major challenges related to interoperability. Adding to the complexity are the challenges related to deciphering all the different machine output files as well as how to standardize into a single format for easy sharing and exporting…. This connectivity conundrum is precisely why global EMS providers are turning to third party partners to access machine data.

Solutions like Co-NECT offer out-of-the-box software interfaces tailor-made for specific machines and host software systems. Running “on the edge” (installed directly on the machine PC, vendor’s line host, or a dedicated interface PC or server.), Co-NECT taps into multiple data sources on each machine (or host platform). With a flexible architecture, Co-NECT also enables simultaneous transmission of data from a single machine to diverse systems and other machines, each employing varying protocols. For machine to system or system to machine communication, Co-NECT utilizes REST APIs for sending and receiving commands, promoting a unified flow of data and information.

Put simply, factory analytics should transcend mere data visualization and dashboards, and envelop a comprehensive solution that Cogiscan's Factory Insights epitomizes. This advanced performance analytics platform, built upon decades of expertise, doesn't just display data; it intelligently harnesses information from across the production ecosystem to reveal and analyze trends, compare performance, and identify hidden inefficiencies. Cogiscan's ongoing commitment to innovation in data management and analytics ensures that their clients are equipped with the right insights to differentiate and excel in this highly competitive environment.