By: Ryan Gamble, President and Founder, Intraratio Corporation

New Product Introduction (NPI) is the highest risk portion of your product life cycle.  Investment cannot be recouped until the product is able to be manufactured in volume, yielding at the right level of performance, reliability and cost.

Exacerbating this risk is the growing complexity and integration of electronics systems in today’s products.

Automotive production is seeing increased reliability issues due to the growing electronics features in today’s vehicles.

Medical device technology is using more highly integrated semiconductor technologies, and field failures can be extremely costly, if not catastrophic.

NPI manufacturing risks fall into three categories:

  1. Line Failures. These are due to assembly, testing, supplier reliability, or product manufacturability issues
  1. On-time Delivery Failures. These are due to manual data collection and tracking, and lack of visibility to product assembly cycle times
  1. Field Failures. These are due to poor product reliability, and/or product performance.

The ability to track product genealogy down to a unit level is critical to addressing these three categories.  Product needs to be tracked through all stages of the assembly process, along with supplier component source and performance data, assembly and test stations used, and the environmental conditions measured while going through various processes.

Unfortunately, most systems in today’s factories still operate as islands, not fully connected to the IT infrastructure, lacking the ability to capture and store the required data for full product traceability.

So the majority of production floor operations use spreadsheets and paper.  This is an expensive and time consuming layer of ‘human glue’, limiting the ability to scale and automate data collection, and instead increasing overall engineering costs and time to market.

One solution is to drive vendors of assembly and test systems to open up their software for expanded data collection and integration with back end databases.  Data collection must be relevant and portable.  No longer can data be locked up on an assembly system, only to be forced to pay additional license fees, and still not be able to access the data via other more powerful analytics tools.  But getting vendors to change is a slow and expensive process, and the urgency is today.

A second solution exists, which is rapidly deployable and scalable.  By quickly automating data collection and product genealogy tracking, this logically leads to the ability to further automate the actual movement of product through the production line.

It involves upfront serialization of the units to be produced, right when they are started.  By mapping the flow of the assembly process, digital ‘signals’ generated by each stage of the process can be identified for use in relationally tracking material, actions and results.

Once serialized, the use of barcode or RFID scanners is easily employed.  This can be coupled with low cost single-board computing tools to automate data capture and processing, so that assembly units are digitally tracked.  Time based signatures can be used to tie units to machine operations, and bind environmental information to the data sets.

This results in the power of real-time statistical monitoring to predict failures, by actively looking for shifts in distributions across assembly, test and even supplier component data.  These same monitors become tools for rapid root cause of field failures, where the focus shifts to tightening of QA test limits, identifying any test escapes, etc.

The backbone to all of this is a strong product genealogy tracking database.  By automating and tagging material through the line, serialized lot and unit traceability enables powerful and robust monitors.

Your NPI risks become well managed.

Production costs go down.  Time-to-deliver is markedly improved.  Field failures are reduced, with resolution times greatly improved.


Addendum: Additional Implementation Notes

  1. Line failure mitigation:
    • Leverage real-time statistical monitors to predict failures
    • Log unit genealogy through serialized lot and device assembly transactions
  1. On-time delivery failure mitigation:
    • Use digital transaction history to track actual cycle times, identify areas of focus for improved throughput
  1. Field failure reduction:
    • Analyze assembly and test data daily, with the goal of identifying performance and quality test limits that need adjusting.
    • Implement outlier detection algorithms on assembly, test and supplier data.
    • Apply automated, random, quality sample testing
    • Look for additive signatures, to quickly root cause the failure, identify other affected material, contain it, and make the necessary changes to resolve it.