Reengineering—sometimes called business process reengineering, or BPR—originated in a 1990 article in the Harvard Business Review by former MIT professor Michael Hammer, who followed up in 1993 with the seminal book Reengineering the Corporation.

The core of reengineering is a change in organizational structure, in which people who were formerly organized by function (“functional silos”) are instead organized by process or product.

That is, all the people who need to work together in a process or product report to the same manager, instead of separate functional managers. There’s more to it than this, but this is the heart of reengineering.

Functional silos. “Functional silos” are at odds with producing a great product quickly and efficiently. For example, a new product may start out as an idea from Marketing, then Marketing “throws it over the wall” to R&D, which develops technologies to meet Marketing’s idea. Then R&D throws it over the wall to Engineering, which designs the product. Then Engineering throws it over the wall to Operations, which figures out how to make it. Once in production, all the people who need to work together to make the product—line workers, technicians, manufacturing and equipment and industrial engineers—report to separate managers. There’s not good flow and communication and coordination between all these different functions. The product suffers. Delays occur. A lot of time and money must be expended just coordinating and communicating between departments, who are working on a variety of products and processes at once.

All the people who need to work together to produce a great product are in the same organization, receiving the same marching orders from the same boss, and are—ideally—in the same room. How could this not lead to a smoother, faster product delivery that better satisfies the customer’s needs?

Often, organizing by product isn’t practical. For example, a job shop that produces a great variety of low-volume custom products cannot have a separate production area or production team for each product (or even product category). When organizing by product at a company level cannot be done, organizing by process can bring many of the benefits of reengineering, even on a scale as small as a particular production area in a factory. That is, everyone working in a particular process—such as a certain production area—reports to a common manager.

The "traditional" organization. I was in a company that made a product that required about twenty production operations. One of those production operations had, working within it:

  1. Equipment operators that reported to the production supervisor.
  2. Equipment technicians that reported to the maintenance manager.
  3. An equipment engineer who reported to the equipment engineering manager.
  4. A manufacturing engineer who reported to the manufacturing engineering manager.

These are classic functional silos. All the people who needed to work together to perform the operation effectively and efficiently reported to a variety of different managers with a variety of different priorities, giving their people a variety of different marching orders. For this reason, problems were very difficult to resolve. It would be hard to get people to attend improvement meetings because they were more beholden to their bosses and their functions than to a particular operation. If people were not doing their jobs, one had to go up to their functional managers to try to get it resolved. The functional manager was often lackadaisical about resolving the problem because this was just one area out of a whole factory that his functional group covered.

The reengineered organization. In contrast, early in my career, I worked at the very successful company Applied Materials, which makes computer-chip-making equipment. Applied had its manufacturing organized into work cells by process, such as Chamber Assembly, Gas-Panel Assembly, and Final Assembly. In each of these work cells were assemblers, a manufacturing engineer, and a buyer/planner, all reporting to that cell’s work-cell manager—organized by process, rather than segregated by function. In my work cell, we would have a weekly process-improvement meeting with all members present. We would discuss problems and ideas for improvement, decide how to implement them, agree on action items, and boom-boom-boom it would all get done. If we had been segregated by function, each attempted improvement would have dragged on for weeks or months if it got done at all. During my work at Applied, the company moved from #7 in its industry to #1, due in part I think to its efficient work-cell-based manufacturing.

The heart of reengineering is replacing functional silos with an organization organized by product or process: all people who work on a particular product or in a particular process report to a common manager.

BOB KOTCHER is president of Simitar Consulting, which specializes in operations improvement using lean Six Sigma methods and computer simulation modeling.

MassChallenge, a not-for-profit accelerator program, aims to “create a bigger pie” by helping innovators get to market.

In a 27,000 sq. ft. office on the 14th floor of a Boston Harbor high rise, scores of innovators in residence toil away, trying to make their dreams reality. Four quadrants of cubicles form the basic layout, abutted by casual open meeting areas, a compact soundproof recording studio, and a galley kitchen.

Read more ...
A 95% first-pass yield is the litmus test. Here’s how to get there.

Capital spending is an internal issue EMS providers constantly grapple with. At the back of the minds of the decision-makers are a host of related and nagging questions. Can we make do with existing equipment?

Read more ...
Do the limits of 2D cameras cast a shadow on their future? by Axel Lindloff

3D inspection has become a buzzword in the industry, but neither the meaning nor the definition of three-dimensional inspection has been standardized. What is behind the buzzword, and what are the different approaches currently offered to electronics manufacturers?
A number of systems and vendors have successfully used 3D measurement for years, but predominantly in the form of solder paste inspection (SPI) or, more correctly, solder paste measurement. It is necessary to differentiate between measurement and inspection, since 3D systems actually measure the solder paste volume deposited on the individual pads. Deploying different measurement technologies such as laser triangulation or phase shift profilometry achieves reasonably good results in performing 3D volumetric measurement, but the task is relatively simple: simply measuring bricks of dull, grey solder paste.

The success of 3D solder paste measurement during the past decade is proven through the process improvements that these systems have contributed to electronics manufacturing, particularly to the paste printing process. The adoption rate of 3D SPI systems in electronics manufacturing has grown from approximately 15% for all SMT lines a decade ago to market coverage of nearly 60% today.

Whether the same measurement technologies used for solder paste can be applied to solder joints depends on the technology. One proven technology in conducting real 3D measurements of solder joints is phase shift profilometry. The task of measuring curved and highly reflective solder joints is complex. The shiny surface and the multi-angled curvature of the solder joint complicate the ability for a laser to map all areas. Similarly, when applying light as the part of the measurement technology regardless of LED or flash, white or multi-colored, specula reflection has to be overcome as an obstacle to obtaining information from certain regions. Multiple directions of light projections must be used in order to compensate for this specula reflection. Too few light directions will not permit such compensation.

In the current debate, 2D AOI technology, which has been deployed for the past 25 years, can’t be overlooked. AOI was the only technology available to replace manual human inspection of assembled PCBs. That there are more than 25 AOI companies worldwide illustrates the need and importance of end-of-line or solder joint inspection in electronics manufacturing. These systems are now active on more than 90% of all SMT lines, and thus much more than that of SPI systems.

Since 2D systems deploy light from multiple angles onto the PCB, components and solder joints, all results for inspection depend on either contrast (e.g., location and polarity measurement) or location of reflection points (solder joint inspection). Figure 1 shows examples of chip resistors with different solder volumes under different light angles.

12kohyoungfigure1a  12kohyoungfigure1b  12kohyoungfigure1c

Figure 1. Chip resistors under (a) low angled light, (b) medium angled light and (c) top angled light.


Even when multiple camera or viewing angles are used, the results are still based on the availability of the applied light. Herein lies the problem; taller components will cast shadows on other components and solder joints. The different lighting conditions resulting from shadowing require the AOI programmer to modify thresholds on a component-by-component basis. As the same type of components on the same board have different thresholds applied, this leads to different defect detection capability and varying amounts of false fails.

Figure 2 shows a typical scenario whereby the same types of leads on the IC will require different settings due to the shadow from the Al-Cap.
Is 3D measurement applicable and really necessary for every aspect of the components and solder joints? Can certain attributes not be inspected in 2D, as they have been for many years?


Figure 2. The tall capacitor casts a shadow on adjacent parts, requiring AOI setting modifications.


The advantages of 3D measurement lie in the word measurement. If you can measure the volume of each and every visible solder joint of every component, why would you just satisfy yourself with counting the bright pixels that the solder curvature is reflecting from an angled light source? Figure 3 shows a comparison of solder joints, measured in terms of their respective solder volumes.




Figure 3. A chip resistor with (a) 15% solder volume and (b) 95% solder volume.


Why discriminate the edges of the component through different colors or grey scales, if the edge can be measured without the influence of component or PCB color in the third dimension? Figure 4 shows a typical scene of a chip resistor, which can be of varying color against the background of a PCB, which can also be of a different hue. Figure 5 shows the 3D image of the same component, where each pixel is measured for its height.


Figure 4. A resistor location shown in 2D.



Figure 5. The same part in 3D.


Currently, AOI systems come with different levels of 3D coverage for components and solder joints. All 3D systems still offer 2D lighting and 2D inspection. It is for the AOI programmer or the QA department to decide to what extent they’d like to deploy new technology for quality assurance.

Contrary to common opinion, the polarity of components is typically a geometric feature rather than just a print. Those chamfers, dimples or corners can easily be measured in the third dimension, rather than relying on varying color and intensity of the print on the surface.  
Should you measure certain aspects of the component in 3D, such as component location, and other attributes in traditional 2D? The solder joint is a three-dimensional structure as well. From a measurement perspective, we think applying the lower level 2D technology to the most critical feature on the PCB, the solder joint, doesn’t make sense. Instead, applying 3D measurement to as many aspects of the components and solder joints on a PCB is the best approach to improving the quality assurance of assembled PCBs, by utilizing true measurement technology with quantifiable measurement thresholds. For example, it is easy to understand that a lead is lifted for 170µm and hence is bad (out of range) if the applied threshold is 150µm. No more interpretation of this result is necessary, no image processing knowledge required and certainly no room for false calls or false accepts. Quality requirements can be translated into AOI thresholds. It makes these values transparent and understandable for everyone involved, e.g., the AOI programmer, the quality department and the operator at the AOI review station.
3D measurement is not just a buzzword, but the future technology for AOI. AS 3D has proven its capability in volumetric solder paste measurement, we would expect the capability to view components and measure solder joints in 3D will prevail in AOI as well.

Axel Lindloff is an applications engineer at Koh Young Technology, Europe;

As the electronics manufacturing industry worldwide has changed dramatically, marketers of OEM capital equipment have changed their strategies and methodologies. Successful marketers have learned to think outside the box – and in some cases, even outside the exhibit hall.

Read more ...

The new PCB flatness guidelines cover machine selection, measuring and testing, to which our author adds some best practices.

Read more ...

Page 33 of 47