Michael Ford

The need for a common language – one written for electronics production – has never been greater.

Manufacturing execution systems (MES) software solutions bring a needed level of management and control to the shop floor, but are often perceived as too expensive for smaller companies. Yes, the purchase price of the software can be high, but the main concern is about the associated costs of operating the MES, including the introduction of checkpoints for material logistics flow, which are manual operations not previously required, and incorporating it in key decision-making processes.

As the PCB industry moves to automate MES applications, such as those suggested by Industry 4.0 and Smart Factory innovations, a highly accurate, timely, and automated data-capture mechanism is needed so these computerized solutions and decision-making processes can be realistic and reliable, without human interpretation. The result will be a much-needed boost of value enhancement and cost reduction for MES operations in electronics manufacturing. The Open Manufacturing Language (OML) is an important first step to kick-start this industry revolution.

An automated process that replaces a manual operation comes with well-known and practiced methodologies for the consideration of operational costs, capabilities, flexibility, etc., all of which are a key part of the return-on-investment study. A key consideration is the flexibility of the automation so that it can quickly adapt to different requirement scenarios, such as the application to a different product or variant. The challenge for automation is the time and effort required to do this. Once set up for that job, however, there should be little or no variation in the way in which the operation is done.

Human flexibility, on the other hand, can have a relatively quick changeover process, for example, because materials need not be set up as precisely as for automation. With human operation comes the higher risk of mistakes, however, especially in the time immediately after changeover and other variation often caused by distraction. But automation has additional requirements compared to the human operator.

The same kind of consideration process must also be made for the automation of decision-making. Decision-making on the shop floor today, beyond the most basic, is dealt with by a human team of engineers and managers who take the available data and together make a judgment. Obvious flaws or omissions in the data have limited consequence because discussion and additional investigation is routinely done before decisions are implemented.

The most common decisions to be made are around execution flow caused by equipment breakdown, significant process bottlenecks, quality issues, changed delivery requirements, material shortages, etc. Effective production management skill and experience continue to be a critical part of the successful shop-floor operation, with MES systems effectively providing operational support. What is really needed now is for these management decisions to be supported by computerized intelligence, which can recognize opportunities for improvement and react far more quickly and accurately. It can calculate the costs and values for each change so they can be executed immediately, at minimal risk.

In this way, optimization of the shop-floor operation, even in high-mix environments common today, can be performed frequently. In doing so, the factory achieves levels of productivity comparable to a high-volume operation, but with the flexibility to react to changing demands and circumstances on a very short-term basis.

The key for the success of this technology is the availability of reliable, timely and accurate information from the shop floor. This is ironically the one thing the electronics manufacturing shop floor lacks. The reasons for this date to the start of SMT manufacturing when SMT machine vendors created interfaces used predominantly for machine tuning, trials and debugging. The various data ports were simply created in a way that was best for the machine, and the data content was determined by the specific technology and operation of the machine. Seeing the need for collection of data from the machines by customers, these interfaces gradually became modified and adopted for external use, mainly in restricted circumstances.

Standards such as GEM-SECS and CAM-X came into the market over time, but were mainly focused on resolving protocol issues, without really addressing normalization of the data content, especially for more recent and complex SMT machines. They also did little or nothing to support non-machine processes such as manual assembly, inspection test, and repair, without which key metrics such as overall equipment effectiveness (OEE) were incomplete. This meant the value that could be derived from the data was limited. Such standards, in spite of significant effort, were not widely adopted in SMT-based electronics manufacturing.

The reality of the situation today is that unless something changes quickly, the electronics manufacturing industry is going to be left behind when it comes to advanced computerization solutions such as those from Industry 4.0 and Smart Factories. What could be even worse is the situation arises where a broader and more generic IoT standard is forced onto electronics manufacturing so it is horizontally integrated with other areas of manufacturing. Instead of bringing value to the shop floor, it actually creates a higher burden, including yet more operations, plus needless investment into new machine platforms and redevelopment of IT solutions.

It would be disappointing if electronics manufacturing were to be put into this situation, at a time when the importance of electronics is accelerating, with electronics-based technology now a part of almost everything we do. Rather than being treated as the lowest common denominator in the industry from the perspective of open communication, it would be far better for electronics manufacturing to take the initiative and lead. The question is how.

The new Open Manufacturing Language (OML) addresses these issues. Unlike previous data communication standards, OML defines not only the architecture and protocol of inter-process communication, but also the data content. OML is bidirectional, so it can be used for manufacturing control and management, as well as data acquisition and visibility. The specification for OML is available without cost, upon registration at A software development kit (SDK) with supporting technology for rapid adoption is also now available.

The OML language represents the opportunity for the electronics manufacturing industry to establish the Internet of Manufacturing using a normalized data flow through which information from any type of machine, from any vendor, can flow, with the addition of data from all manual operations and processes. The scope includes the whole of the shop-floor operation, providing a complete picture with complete control. With accurate, real-time data available through OML, computerizations such as Industry 4.0 solutions can be created in a practical way that yield significant operational benefits, performing key optimization decisions with negligible risk. Such applications include:

  • Visibility of the live production floor, with KPI metrics used to quarantine “out of control” processes through poka-yoke.
  • Dynamic closed-loop line parameter adjustment.
  • Automated Lean material logistics.
  • Production flow analysis and optimization.
  • Finite production planning.

The values that can be associated with the use of OML scale from the simplest step-by-step, low-cost, no-risk approach for individual solutions by the industry's smallest companies, to the Tier 1 multinational manufacturers that need global analysis of operations at their fingertips. OML can be used to optimize current ways of operation or to assist with the introduction of new operational methods. There is no prescriptive flow, and every application is able to provide whatever is right for the operation.

As OML finds a home within electronics manufacturing, delivering values with short returns on investment, OML data can then integrate with broader, less-detailed IoT standards so there is a hierarchy of value for IoT with appropriate costs. To make a single IoT standard able to support the level of detail needed to create value from Industry 4.0 computerizations across all industries is unlikely to happen because the data content specification would become very complex, verbose and often conflicting. Having a fully detailed solution focused on an appropriate scope within the industry allows the full value and opportunity to be realized.
Broader horizontal strategies are left to focus on common data elements appropriate for higher levels of management and control.

Here is the challenge, not only for MES and smart factories like Industry 4.0, but also for the development of IoT. Will there be similar formats for other industries?

Ed.: Mentor Graphics conceived and developed OML.

Michael Ford is marketing development manager, Mentor Graphics (; His column runs bimonthly.

Register now for PCB West, the Silicon Valley's largest trade show for the printed circuit industry, taking place Sept. 13-15 in Santa Clara: PCBWEST.COM

Submit to FacebookSubmit to Google PlusSubmit to TwitterSubmit to LinkedInPrint Article
Don't have an account yet? Register Now!

Sign in to your account