Electronics Assembly .… Exposed

Exposing PCB Manufacturing, Assembly, and Test processes to improve efficiencies, productivity and business!

4 January, 2016

In this second installment in the series of “Leveraging an Analytics Strategy”, we discuss the prerequisites and data sources.

Creating value that can be monetized is at the core of leveraging manufacturing analytics. Analytics is still a product of “good data collection,” but with the end value in mind. Deciding which data should be correlated and what analytic models to use should be done carefully, because as the figure below describes below, “garbage in, garbage out,” in both the quality of the data and the analytic model.


Technology Confusion

When discussing analytics today, there is a natural tendency to associate this topic with just the infrastructure of “Big Data.” Big-data infrastructure is used by retailers to track user web clicks to identify behavioral trends that improve campaigns, pricing, and inventory. Utility companies use it to capture household energy-usage levels to predict outages and to invent more efficient energy consumption. Governments are using it to try to detect and track the emergence of disease outbreaks via social-media signals. Big data is new and unique in collecting large amounts of data to be processed quickly with complex correlations.

Databases such as Apache™ Hadoop® or other NoSQL databases with analytics infrastructure such as MicroStrategy™ and others are big-data technologies. After the volume of data is reviewed, there are many options. Traditional databases such as Microsoft’s SQL Server, Oracle, and others may still be just as effective in leveraging manufacturing analytics for competitive strategies. Let’s not confuse the technology with the goals.

Data Sources

Achieving the level of analytics needed would require several different, and currently disparate, systems to come together into an analytics engine. Those systems typically would include:

  • Enterprise resource planning (ERP)
  • Design for manufacturing
  • Manufacturing execution system (MES) and/or quality management system (QMS)
  • Supply-chain integration software

Most ERP systems include the accounting and purchasing of materials and, in many cases, also maintain the inventory and high-level scheduling. From ERP, information about purchasing trends of components, availability, lead-time plans vs. actuals, obsolescence, and, most importantly, cost would be available. Those may be cross-referenced to the Bill of Materials of projects based on products or programs, to get real trends and measurements.

A DFM solution would provide risks, cost avoidance opportunities, and constraints adherence. Issues that may affect the manufacturability of a design would be identified early, as shown in Figure 3. Manufacturers are able to price their services more accurately and can show cost-drivers in a more granular way. For example, if DFM analytics show that a certain design issue is likely to decrease yields by 2%, then price can be adjusted accordingly.

Line utilization, machine performance, feeder performance, OEE, first pass yield, RTY information tied to the stages in manufacturing, then to design information, components, materials, customers, and revision should be accessed and fed into the manufacturing analytics engine. Powerful correlation data is available when integrated, for the use of building new business models by manufacturers. This is the most complex integration, but it also provides the highest potential value.

Finally, integrating supply-chain sources into the manufacturing analytics would be ideal. Several data aggregators in the industry provide this information through Web-based connectivity tools for automatic query and retrieve capabilities. IHS CAPS Universe™ and SiliconExpert Technologies are a couple of component data providers, although there are several more. They provide a lot of detailed data, including:

  • Electrical part parametric data
  • Part obsolescence
  • Part Change Notification (PCN)
  • RoHS/compliance
  • Market part availability
  • Lead time (from vendor partners)
  • Inventory information (from vendor partners)
  • Cost

Also, the Electronic Components Industry Association (ECIA) maintains a list of partner companies

Next installment will include more specific examples of how manufacturers are leveraging analytics.

Looking forward to getting feedback on how some of you leverage analytics in your environments!

, , , , , , , , , , , , , , , , ,

23 November, 2015

Business Intelligence, Data Analytics and Big Data are the hot buzz words in the IT Infrastructure arena these days. In our world, every manufacturer collects data anyway.  They may be collecting data based on customer requirements, compliancy requirements, standards requirements (e.g. ISO), internal metrics for more informed decision making, or for sake of just collecting data. In many cases this data is difficult to access, it is in multiple sources, and the true decision making and organizational tactical management is done with spreadsheets and whiteboards. How to really benefit from the collected data and convert data into “actionable information” is the work of an analytics strategy.  With the right structure analytics can improve operational efficiency and improve visibility for more effective decision making. These metrics are the key to attracting and retaining customers. In addition, manufacturing analytics can become a powerful weapon and a competitive advantage for additional revenue streams.

Enabling manufacturers to develop services around the analytics “information” may be the key to improving partnerships and long-term business. Design organizations are looking for partners in manufacturing to not only build “on time”, “on budget” and “at quality standards”, but to improve their designs and products. What if a manufacturer can provide “information” to enable their customers who are looking to improve their component selection process, ideally choosing parts with the lowest defect rates and lowest counterfeit rating? What if a manufacturer could provide DFx “information” about every revision of a product, how it compared to the previous revisions (not only the last revision) to ensure violations and waivers were monitored and ensure corrective actions were truly done, and old violations were not re-introduced in a future revision? How much would design customers value a dashboard with production status, Roll Throughput Yield, and running defect correlation to design best practices for their products? What if a manufacturer could provide a “service” to monitor quality metrics per product or program further strengthening collaboration with the design customers and allowing them key insight to how designs can improve over time, further leading to improved customer service and customer retention? There may be opportunities to make additional services revenue leveraging Manufacturing Analytics “information” as well.

This is a kickoff on a series related to using using analytics as a methodology for business development and customer retention.

Looking forward to hearing thoughts and feedback on this topic, as we go through the various customer stories and scenarios.

, , , , , , , , , , , , , , , , ,

22 September, 2014

Have you underestimated the costs of manufacturing a product?   Cost overruns killing your planned margins? Don’t really know if design decisions have any effect on final manufacturing costs?  Under pressure to reduce your costs, but don’t understand where the costs are coming from?

Some say understanding the “cost of manufacturing” is an elusive number.  However, with a well-designed systematized Factory Model, simulating manufacturing costs IS achievable.  The factory model should support labor rates for different operations and roles, various overhead costs, built-in margins, etc.  The simulator should be able to calculate Bill of Material line-item costing.  In addition, the design data itself must be analyzed, such that Design for Manufacturability parameters are extracted and used to further fine-tune the manufacturing cost simulation.

For product design organizations struggling to ensure designed-in costs are in line with final expected costs, a Factory Model based manufacturing cost simulation would be vital to ensure design decisions are made for optimal balance of cost and function, or assist with creating baseline costing for product planning.  Feedback during the design cycle would dramatically limit potential added costs, knowing that 70% of all costs for a product are expended in the design stage.   In the hands of contract manufacturers, there is no better way to ensure responses to RFQs are handled so they can “quote to build, build to quote”.  Margin points, profit or loss can be determined by variables that were not considered.

Although there may seem to be an infinite possibility of variables, creating a systematized factory model based manufacturing cost simulator with integrated bill of materials line-item costing and DFM analysis, will produce a closer-to-reality simulation and dramatically improve margin points and decrease unexpected product costs.

Join me for a presentation and discussion at SMTA International 

SMTA International

Spotlight presentation in the main exhibition hall, Wednesday, October 1  from 11:00am – 11:30am.

Hope to see you there.

, , , , , , , , , , , , ,

17 June, 2014

Tired of entering the same recommendations and descriptions for design DFM issues found over and over again!!! You are not alone.

If you are one of the hundreds out there that are responsible for providing feedback for PCB Design issues, read on!

As the manufacturing expertise moves away from the design organizations and into the manufacturing organizations, effective collaboration tools are needed to ensure communication flow is not disrupted.  Electronic Contact manufacturers are getting involved with design layout decisions earlier and earlier in the product design flow.  Outsourcing design layout is further complicating the collaboration and feedback loop between manufacturing know-how and the designers.  Identifying potential fabrication, assembly, and test problems in the PCB design before moving to manufacturing can offer huge benefits.  Many are also using the manufacturing expertise to provide DFx Services to design organization.

DFM report

Automated reporting

Consider automation!

With new advances in the Valor DFM tools and the updated Manufacturing Risk Assessment functionality, it is now easier than ever before to tag issues, add comments, and export results.

Automation has been developed to automate the process of generating these reports. In most cases, reports that once took hours to create can be done in minutes. This automation provides Valor NPI users a quick and easy method for creating professional looking, consistent DFx reports which are configured specifically for each of their customers.

Auto-populating of Recommendations, Reference Documents (e.g. IPC 7351) based on the category name, componrecommendedent type, etc.. saves hundreds of hours of manual editing.

In addition, collection of design metrics would assist understanding design complexity and risk assessment (DPMO).

Tired of entering the same recommendations and descriptions for design DFM issues found over and over again!!! You are not alone.

Contact me if you want to discuss how to make this happen for you or your supplier!

What are your main challenges with DFx collaboration?  Is the market ready for moving from static reports to something more?

Looking forward to any and all comments!

30 May, 2014

Are mobile Apps becoming more prevalent in the Electronics Manufacturing landscape?

I recently came across a survey at Manufacturing Business Technology magazine by Canvas:


The article states “…The survey points to an increasingly mobile manufacturing workforce that is looking to extend more day-to-day business processes to their mobile devices”.  The key areas of mobile device usage seem to be in the inspection of goods areas (48%).  Although this survey is more generic in nature and not specific to electronics manufacturing, I would argue that Quality Inspection, Inventory Management, WIP Management, and Test areas are ideal candidates to deploy mobile app devices.  Locations where data collection is vital to maintaining visibility, aggregating metrics, and feeding manufacturing operational intelligence should be areas of focus.  In addition, real time feedback to the end-user, such that critical quality and management decisions can be made, in-process quality may be implemented and a dramatic reduction of paperwork, manual entry can be achieved are additional reasons these specific areas of focus should be addressed by all electronic manufacturers.  In addition to Valor’s MSS technology that can provide these capabilities, factories are many times looking for internal resources to develop such capability.  Thorough analysis of cost of ownership should be done before deciding on approaching vendors or developing internally.  Are there many electronics manufacturers implementing mobile apps and devices on their factory floor?  Yes!  However, this push seems to be more prevalent in Europe than in the US right now.

So, how does this fit into the Internet of Things?

Internet of Things (commonly abbreviated as IoT) is used to denote advanced connectivity of devices, systems and services that goes beyond the traditional machine-to-machine (M2M) and covers a variety of protocols, domains and applications. – defined by Wikipedia.  I’ve been thinking about how the mobile manufacturing workforce may be extended so that more connected devices are deployed and all types of equipment may feed data-aggregation strategies of business intelligence to drive manufacturing excellence (at Cost, in Quality, on Time).  Some examples may include equipment that order material on its own, when levels are low, or kanban levels that drive shop orders for refill.  With the expected upcoming announcement by Apple of their MFi Certification project, I believe there will be more push to have connectivity between ‘wares’, and that should creep into the manufacturing space as well.

I’m sure there are others who believe as I do, that it’s only a matter of time before the value of business intelligence and push for manufacturing excellence will drive more mobility and connectivity in the competitive manufacturing workspace.

I’m eager to hear your thoughts and ideas!

, , , , , , , ,

5 November, 2013

What if you had a tool that could prevent problems in your processes and products before they occur?

Failure Mode and Effect Analysis (FMEA) is that tool.  FMEA is a formal process of identifying all the ways a process or product can fail, and then determining how to reduce or eliminate them.

Product development groups and quality groups work tirelessly to identify ways their products and processes may fail.  Identifying the ways products or processes may fail is the first step in either designing-in improvements, or taking steps to ensure those failures do not occur.  There are several structured methodologies that can be used to work through defining potential failures and working through their effects.  Failure Mode and Effect Analysis (FMEA) is one of these structured methodologies.  It provides a way to examine the problem area, identify potential failures, identify the effects of those failures, and quantify those effects through various ranking values to come up with a priority rating that will allow teams to focus on the most critical potential failures, or failure modes.  Although initially developed by the military, FMEA methodology is now extensively used in a variety of industries.  The method is now supported by the American Society for Quality which provides detailed guides on applying the method.

There are several steps to using FMEA:

Scope of Study (Area to focus) – Defining the scope of a Failure Mode and Effect Analysis is very important.  It must be well defined, with clear boundaries for the organization (management) and team members.

Identify Possible Failure Modes – The goal in this step is to identify all the ways things can go wrong.  It is important not to get wrapped up around what that effects of the failure are…that comes next

Identify Possible Effects of Failure – Now that all the ways things can go wrong or may fail have been listed, each failure mode should be evaluated for the effects of the failure.

Define Severity Ranking – Severity ranking, as with Occurrence ranking and Detection ranking, are based on a 10-point scale, with one (1) being the lowest ranking and ten (10) being the highest ranking.  Lower rank means a less severe failure mode. Typically in a Design FMEA, no effect would get a severity ranking of one, whereas noncompliance to regulatory requirements, or may cause harm to customers would be assigned a severity rank around 10.

Identify Occurrence Ranking – How often will a given failure mode occur?

Identify Cause of Failure – Brainstorming the cause of each failure should be done at this stage. Typically, subject matter experts, equipment vendors, vendor consultants, or outside consultants are used to fully analyze the causes of the failures.

Identify Detection Ranking – Detection ranking looks at how likely is it to detect a failure or the effect of a failure.  Analysis of current controls to the product development flow should be done to determine the likelihood of detection.

Calculate Risk Priority Number – The Risk Priority Number defines the overall risk of a failure and its effect to the process or product.  The Risk Priority Number (RPN) is calculated with a simple formula:

Risk Priority Number = Severity x Occurrence x Detection

The total risk priority number is calculated by adding the individual risk priority numbers of each failure and effect. Although this number in itself has no meaning, it is a way to gauge RPNs for a specific FMEA study as improvements are being made.

Prioritize Corrective Actions – The failure modes can now be prioritized by ranking them in order of the highest risk priority number to the lowest.

Now that we know how it works, how can we use it?  One of the initial steps of an FMEA study is to identify possible “Failure Mode Causes”.

DFM Solutions as part of FMEA

DFM Solutions as part of FMEA

Identifying ways products can fail in manufacturing is a tedious and often times trial and error process.  Knowledge from years of previous product development and realization processes culminating in “out of the box” set of possible failure mode causes would dramatically improve identification of areas of risks, and effects of failures. Working through the FMEA processes outlined earlier, most DFM solutions come with analysis tools to identify potential design for manufacturability issues.  As DFM manufacturing solutions are implemented, they can serve as both detection and control mechanisms for the FMEA process.

One of the most obvious examples is the potential failure mode caused by copper flooding under components with multiple ground connections, or similar condition.  When assembly manufacturing PCBs which have passive components with uneven copper distribution between their leads, the differential in heat dissipation rates between the two leads may potentially cause cold solder joints and or slightly weaker solder adhesion, and thus leading to a potential tombstone problem.

In addition to aiding in identifying failure mode causes, DFM solutions may play a large part in providing control and managing the detection of the failures, before they happen.  Detection should be managed with dual Design and Process control. This interrelationship between Design and Process should be maximized between Design FMEAs and process controls in place in manufacturing with parallel Process FMEAs in place.

I realize this is an overly simplified overview of a complex and comprehensive quality tool.  However, I’m hoping the realization that good quality tools are out there, and design organization need not look any further than their manufacturing partners, vendors and suppliers alike for best practices and ways to identify and improve product quality.

Looking forward to any comments on the this topic, and FMEA’s applicability within the Design space.

Mentor Consulting provides customers with expertise in electronic design and manufacturing infrastructure and methodology. Valor Services include a worldwide team of professionals including former factory managers, engineering managers, software developers, designers and various experts who have improved PCB Assembly factories globally.  Mentor Consulting solutions are engaged worldwide by forward-looking electronics companies to optimize design and manufacturing productivity and advance adoption of the latest industry best practices. For more information, contact mentor_consulting@mentor.com.

, , , , , , , , , , ,

14 December, 2011

Production Schedules….Scheduled! – “General Guidelines”

In this post, I’d like to review some basic guidelines that most production planners and schedulers seem to agree upon.  Not meant to be the cure-all defacto standard, but based on polling several factories of various sizes and strategy seems to be a core list.

1)     Plan the schedule to ensure up to 99% capacity of the bottleneck operation or resource.

Planners should plan for the bottleneck operation or resource to be constantly and smoothly worked by providing and scheduling material and work for that bottleneck operation or resource.  This may also require larger buffers in front of these specific operations or resources to ensure smooth, constant work even if the upstream operation is starved or goes down for a period of time.  In some factories, burn-in is a bottleneck, while others it might be test or even specific equipment in an operation such as the reflow oven.  If the bottleneck operation or resource stops, there is no chance of hitting the-on time delivery metric.

2)     Align scheduling and planning (work order release) to business goals (on time delivery vs. maximum throughput vs. minimum cost or minimum WIP).  This might vary from quarter to quarter.  However these trade-offs must be considered in the scheduling.

3)     Agree on the metrics and checklists for feasible lead time quoting to Sales.  When sales promises undeliverable timelines, no one wins.

4)     Calculate “lead time variance” and set start dates based on the lead time variance to achieve delivery goals.

Chart distribution of lead time vs. the probability of achieving that lead time.  If the goal of on-time delivery is 95%, then we need to plan job release using a lead time to fall within the 95% lead time probability.  This might practically be eight weeks out, although the average lead time in a factory may be six weeks.  If a planner releases jobs and plans on the average lead time, then by definition the plan will miss 50% of the job’s on-time delivery measurement!  The average calculation determines that 50% will be less than that average value, and 50% above. Instead of relying on the average, plan for the practical job release to hit on-time delivery.

5)     Calculate “quality variance” to plan overage

Calculate yield distributions vs. the percentage of expected overages for an order to meet delivery requirements.  Using this framework, a production planner can use a more reliable metric to calculate the overage needed to produce the required quantities considering quality fall out.  However, if you plan for too low quality, then the overage amounts will be too high, and lead to waste.  The same framework can be used to schedule safety stocks of semi-finished goods.  For high-runners or frequently re-ordered products, only the amount to cover forecasted uncertainty should be calculated and kept in stock, rather than building overage each time.

6)     Use real-time information on assets, resources, material, WIP, yield, etc. to ensure a feasible daily job release and scheduling on factory floor.

7)     Use real-time notification that the schedule can no longer be met.

Production planning and factory management should be notified if schedules can no longer be met due to factory condition changes.  These may include yield, machine availability, productivity levels, and resource availability.  Manufacturing Execution Systems or Manufacturing Operation Management systems (e.g. Valor MSS™ ) would be key to achieving this goal.

8)     Schedule must be available to everyone, and easily understood.

Publish the schedule and have it available for everyone to see.  This ensure visibility to the risks and goals in front of every department.

The above guidelines are a consolidation of discussions by various factory planners, production planners, and ERP users, in addition to references selected.

What of the above is in practice in your factory?  What is not, and should be?

Next installment we cover other factors that affect production schedules.

Stay tuned…..

, , , , , , , , , , ,

1 November, 2011

Production Schedules….Scheduled!

“Strategic Alignment”

There are many variables the influence meeting production schedules.  One of which is factory strategic alignment.

The goals of a factory are heavily influenced by the incentive programs deployed in the factory.  These incentives should be driven by that organization’s competitive strategy in their markets.  Are they focused on advanced technology, low-cost producer, highest quality producer, fastest turnaround, or on-time delivery and customer satisfaction?  These incentives are, in most cases, conflicting strategies, and an organization can only prefer one to two of those.

Are the organization’s individual department incentives aligned with the factory goals?  If Production is incentivized by total throughput (total number of products shipped), there is a risk to quality and of using line capacity for Engineering process improvements or changes.  If there is an incentive for quality, throughput and engineering test/process tests are at risk.  If Production is incentivized for throughput, but Planners are incentivized for on-schedule ship, and the Test department is incentivized only on First Pass Yield or Final Yield, etc., you have conflicting incentives between each department causing chaos and an increased probability that the actual schedule will be ignored.   Each department tries to achieve their own goals and the overall goal of improved on time delivery, on quantity, on cost will not be achievable because of departmental priorities, use of equipment decisions, priorities in schedules, and determination of  individual measures.  Each of these criteria is a shared responsibility with all departments within an electronics manufacturing plant.

Also, with the above variable incentives, it is quite easy to “blame” departments for specific problems when dealing with delivery challenges.  The most common assignments of departmental “blame” seem to be:

  • Sales accepted an order with less than the required manufacturing lead time
  • Purchasing didn’t provide the parts on time
  • Sales agreed to an order without checking into part lead times
  • Production focused on running the high-runners rather than getting the smaller jobs through
  • Production operator errors keep making schedules slip
  • Equipment breakdowns are making schedules slip
  • Production planners have no idea what’s really going on in the factory
  • Process variation due to engineering processes causing scrap and rework
  • etc…

All managers in their departments should be measured on common measurements, aligned with factory strategy, in addition to any department specific measurements.  Achieving these measurements is a group effort, and not only any single department’s effort. When all departments are evaluated, they must work together to achieve the measurement goals and thus more likely to meet the production schedules.

Are the departments’ in your factory incentivized to meet production schedules and aligned with each other?  Are they aligned to the company strategy (even if it changes every 6 months)?  What measurements do you use to know how well production schedules are met?

Next installment we cover other factors that affect production schedules.

Stay tuned…..

, , , , , , , , , , ,

20 October, 2011

Production planning is a dynamic problem to solve, as the variables seem infinite.

As more and more global regions move to high-mix with increasing volume production, the planning task of running multiple products effectively through existing assets and resources poses a significant scheduling challenge. Manufacturing plants today are challenged with shipping the required quantity of products on time, at quality levels, and at planned cost. Over the years there have been many automated systems developed to “simulate” the production environment and try to lock down many of the variables so that schedules can be “optimized”.  Although there is a large diversity in opinions whether those approaches work, there is commonality in goals and factors to consider. Tactical production planning is currently done with tools such as MS Excel or even a large whiteboard on the factory floor.

Many variables affect a schedule. Is there a system in place that can provide the information needed to put together a “good” schedule, then measure it and improve moving forward? As we move from a monthly schedule, to weekly schedule, then to a daily schedule, these variables have a large effect on the schedule’s success.

The goal should be a consistent schedule whereby products are shipped on time, at planned quantity, at planned quality at planned cost.

Let’s explore the best practices, influences, challenges, practical measurements and metrics and finally tools available to meet those challenges.

Stay Tuned….

, , , , , , , , , , ,

9 February, 2011

Still chasing the dream to achieving optimum manufacturing flow?

We have arrived at the 4th and final installment of the series of achieving optimum manufacturing flow.

The previous posts covered Process Engineering, production planning (Scheduling) and material management in the factory and how we can optimize each of the tasks? 

In this blog post, we focus on SMT and the shopfloor.

Bottleneck management in SMT

Overall Equipment Effectiveness (OEE) is a metric that includes throughput, availability and quality.  It is a metric that allows organizations to view performance of assets in the factory. 

  • Availability = operating time/planned production time
  • Throughput (run rate) = (total pieces/operating time) / planned cycle time
  • Quality = TPY = (good pieces / total pieces)

Based on an Aberdeen Group study, top performing organizations are focused on OEE.  Although one would think that improvement in performance might come with increased cost, the information in Table 1 shows that for the top 20% they are able to reduce annual maintenance cost by 8% and overachieve on their Return on Asset (RoA) by 11%.  It is clear that in order to realize these benefits, software tools such as Mentor’s MSS suite and optimized processes need to be employed.


Aberdeen Study - OEE

Aberdeen Study - OEE

 Since statistically over 80% of components are placed with automated SMT equipment, we move our focus on those specific assets.  There are 3 main elements to optimizing process performance in SMT equipment, and ensuring high OEE:
1) Ensuring maximum availability of assets by minimizing downtime
2) Ensuring high throughput by ensuring optimal machine efficiency
3) Ensuring high quality by monitoring placement quality real-time.


Low Level Warning

Predictive parts-out monitor

 Downtime at an SMT machine typically occurs due to material shortage.  It is not common that SMT equipment is down due to mechanical failures.  One of the most painful bottlenecks in production lines is parts exhaust or parts out.  What happens then?  Typically the SMT Machine alerts the operator, then operator runs to a sub-store or stock area, pulls the material (or worse orders the material and waits), while the machine sits idle.  A software system that has a SCADA level direct interface to the SMT equipment, which is monitoring every placement and ensuring adequate parts are available in real-time can also monitor when material will be exhausted from a feeder.   

This predictive functionality like what is available in Mentor’s MSS vManage product, would alert the line operator that they must be ready with kitted material, so as to minimize the downtime of the machine.  Having a real-time predictive parts exhaust monitor, and regular preventative maintenance programs should ensure high SMT uptime.

Throughput and machine efficiency should also be monitored to ensure that production rates are monitored against planned rates.  Production schedules are based on planned rates of production through the assets.  If the planned rates are unrealistically lower, then production schedules will always be behind.  The key to knowing is having visibility to the actual values.  Actual production rates, many times calculated with the CPH (components per hour) value, should be visible using a real-time production monitoring system.   

Real-time Feeder Errors

Real-time Feeder Errors Monitor

A production monitoring system will allow identification of throughput values, target vs. actual analysis, and alert related production personnel when the rates are unacceptably low.  Typically excessive machine errors due to intermittent component picks lengthen total placement time of a specific board, adversely affecting throughput. Nozzle errors due to vacuum errors or unclean nozzles are one of the causes.  Placement identification errors are another cause of a reduction in throughput.  This is usually on NPI runs, or production part replacement with alternate parts defined in the BOM (bill of materials).  When an alternate part is used, there is risk that a different component vendor may be used which provides the same functionality chip.  However, that chip may physically be a little different in height, lead-length or body width.  AVL Validation tools available in Mentor’s vPlan application can remove this risk.

Finally, ensuring that the quality of the placements is monitored in real-time so excessive pick-rejects are caught before too many parts are rejected. This would allow line operators to stop a machine before component attrition gets too high, further adding to production costs, and troubleshoot the reason for the high errors.

We have ensured high availability and we have analyzed how to ensure good throughput.  Now we must monitor quality.  Ensuring high quality is done by integrating real-time quality statistics to the OEE metric.  Typically an in-process quality step is added at the end of the SMT line to collect this information.  This can be a visual inspection station where a quality inspector is reviewing each board, or an automated inspection system such as an AOI (automated optical inspection) machine.  The Throughput Yield (TPY), also called First Pass Yield, is collected.  A simple approach to TPY can be employed by calculating the number of acceptable units divided by the total number of units produced thus far.  In the real-time calculation reworked units are normally not considered, but in the final TPY calculation, they should.  Mentor’s vCheck system performs real-time calculations of TPY and can feed this information to an OEE monitor in real-time, thus providing the organization a true OEE in a real-time dashboard.  With this visibility of the OEE metric, production and quality engineers are better equipped to analyze and act to optimize process performance.

 On our journey through production planning and scheduling, process engineering, material management and into the SMT area of the shop floor, we’ve seen how systems can dramatically affect manufacturing flow.  Optimizing material flow reduces non-value added tasks, reduces potential areas of production bottlenecks and stoppages, increases the utilization and efficiency of assets, and helps schedule more efficiently with real-time production information, while ensuring material is available where it is needed, when it is needed.  Addressing these areas and bottlenecks can bring you increased manufacturing velocity, leading to an optimized manufacturing factory.

Still think achieving optimum manufacturing flow is a dream?


Looking forward to your comments either here or in our Community site!

, , , , , , , , ,