Beyond software: How process improvement unlocks real manufacturing efficiency

Over the past few weeks, I have had the opportunity to discuss the Siemens Process Preparation solution with many customers and as a software solution to optimize the delivery of design data into PCB manufacturing. The focus of these discussions is how they can best adopt a software solution but as the discussions get into the details of how the new software platform can help them, additional issues are raised as I dig deeper and deeper in the conversations.

One area that consistently comes up in the process of merging a bill of materials (BOM) file with the source CAD to get the as built components into the layout. However, as I discuss this more, I find that a lot of companies don’t have this dialed in and perform a lot of manual manipulation to get accurate data on all the components, including the alternate parts with matching manufacturer names and manufacturer part numbers or MPNs. This becomes a great example of not just adopting a software solution but looking at how process improvement can be considered too to provide a step function increase in value to the organization. Within the Siemens Process Preparation solution, we have a very extensive Software Development Kit or SDK, that can be used to extend the application to address customer specific challenges. Creating simple scripts that automate the manual processes that were previously necessary delivers a faster, leaner and error-free solution that takes the software application to a new level, delivering a significant process improvement than was possible before. Most of the capability is still delivered as a supported third-party solution but additional solutions deliver even greater value.

Moving into the area of test programming, both in-circuit test (ICT) and flying probe test (FPT) can suffer from correct supply of electrical parameters. Having several solutions to automate the addition of these parameters, can address the single longest time contributor to the off-line electrical test programming process. I see many customers spending a lot of time editing various files to then splice this into the BOM file to be able to import it and make it usable in their existing programming process. Like the scenario above, this is error-prone and takes time. With Siemens Process Preparation solution provides many ways to obtain this data so out of the box, it is likely that one of these can be used to reduce the single biggest time contributor in off-line test programming. Here’s some examples of how the application can automate this process.

• BOM import using specific values
• BOM of a description that can then automatically extract electrical parameters from within it
• Extraction of the source CAD files design tool attributes
  • This can be electrical values and tolerance directly
  • It can also be extracting a description from the CAD and then using it to further extract the data from within it
• Accessing another business system that contains the necessary fields
• Running algorithms that can identify the component types and pin types if necessary
  • The latter would be for diodes, zeners, transistors, FETs, etc

This raises another issue along these same lines, which is how diversified the environments that customers operate in are, they must deal with a myriad of different data formats and quality that can mean the difference between getting a product ready for production in minutes or hours or days. That means the solutions they use need to be flexible, robust and learn the most they are used.

The central core of Siemens Process Preparation is to create a single, complete, accurate digital twin of the product to be built. So many customers I talk with are creating a product model in each of the applications they use that each program a specific machine. When they take those programs to the line there is a high probability that they will not be consistent and therefore will not deliver in the goals for manufacturing. The only process that can keep up with smaller and smaller lot sizes and higher and higher product mixes is the singular digital twin. Import one set of design data, import a single BOM, bring in manufacturer specific part data for all components and alternates, then you have a robust digital twin that can supply all areas of the production line from a validated data model.

I have discussed the numerous advantages of using intelligent CAD data to start this digital product model but there are many cases where this does not exist. Aerospace and defense are examples that need to build products that are decades old and only limited design data exists. Having the flexibility to create a product model from minimal data or better yet, using incomplete data files to build up a digital product model far quicker than using pure unintelligent data can save hours in preparation time as well as making the process much more accurate.

I will continue to discuss our solutions with existing and new customers; I am always learning from them as I hope they learn from my experience. Please reach out and let’s talk about how we can help each with process improvement and not just software adoption.