Moving from Reactive to Proactive Traceability in Manufacturing

Traditionally traceability has been a reaction to the requirements of a customer or to a regulatory requirement.  If traceability in manufacturing is to offer real value, and not cost, then it needs to become proactive, creating traceability data as a byproduct of a data driven manufacturing excellence strategy.

The typical tools and systems to meet the traceability challenge of the past were most often adopted out of grudging necessity placed upon the manufacturer by the market, regulatory agencies, or customers.  As such, the approach was often one of ‘do as little as will meet the requirement’.

Consequently, systems were bought or built that would satisfy only the specific scope of traceability demanded, and in many cases only on the particular assemblies or products that required it.  This approach answered the immediate need of the company, but failed to achieve anything further.

Lost Opportunities

This narrow approach is fraught with missed opportunity.  By adopting such a narrow scope and application of traceability, the entire manufacturing culture is trained to see traceability and data acquisition and management as a burden rather than part of the way business is done with inherent benefits.  Furthermore, the solution built has to be customized and extended every time the requirements evolve, incurring additional costs and management time.

The biggest lost opportunity is process and manufacturing improvement.  Consider the scope and depth of data available.  That information can really drive operational excellence when properly utilized.  It can flow back to R&D to improve future designs and it can drive real-time improvement by giving process engineers, operators, and managers a view into the reality of the process.

The data set included in modern traceability effectively becomes the definition of ‘Manufacturing Big Data’.  When that data is leveraged for more than just answering the traceability challenge, it becomes the most powerful process and product improvement tool imaginable.  Once manufacturing big data is harnessed, traceability becomes a natural byproduct.  More importantly, the analytical and process improvements that come from leveraging this Big Data are gateways to Operational Excellence.

The Big Data Approach to Traceability

Manufacturing Big Data is a term used increasingly in manufacturing but rarely discussed at a tangible level.  The enormity of the concept has to be grasped before discussing applications that offer operational excellence and traceability.

Firstly consider the scope of manufacturing operations from when R&D finishes the CAD design and the bill of materials (BOM) is locked down in PLM and/or ERP.  This is the moment at which the readiness of the product configuration is achieved.  There is a good deal of data involved in the mapping of that BOM to the associated CAD and revision data.

The process, quality, test, industrial, and manufacturing engineers along with planners then go to work on that data to transform the design into a detailed process complete with robotic assembly, inspection, test and process machinery programs.  The quality plan and route control logic must be developed and laid in.  The visual assembly instructions for every station of the process flow must be designed and digitized along the process.  This New Product Introduction (NPI) process yields a lot of data mapped to the CAD design and the BOM including revisions, people involved and the work output they created.

We’ve not even reached the factory floor itself and there is already a large body of ‘Manufacturing Big Data’ to store and eventually mine.

So the product is ready to launch into manufacturing and the lines are ready to start?  Not yet.  This is the ‘value add’ axis of operations, and without materials being in the right place at the right time, manufacturing cannot begin.  When the production schedule initiates a build, a hand-off from ERP materials management in the warehouse to the more granular shop-floor control provided by the Manufacturing Operations Management system must begin.

Organization of materials into transport orders targeted to the proper process point, the management of local stores, kanbans, and even the interception of materials into delivery or feeder systems and the exact point of use are all needed.  All of those functions must be traced down to the part, lot, and delivery package so that materials traceability has a record from the incoming loading dock to the point where the material or part is consumed into the product or process.  This critical engine of operations must run constantly in the background, feeding the flow of materials to production, and the flow of unused parts back into the warehouse all with precise tracking.  These activities alone yield vast, valuable and critical data sets within the ‘Manufacturing Big Data’.

With materials present and correct at each station, the manufacturing process itself can begin with the serialization range and work order information flowing into operations from ERP.  The flow of the product begins, and data starts flowing in from conveyors, the assembly and process data feeds, the operator actions and confirmed presences at each station, the materials, chemicals and tools consumed or utilized as a product entered each station.  Vast amounts of information can be gathered about every conceivable environmental variable surrounding, as well as any entity or material that was added to the product in addition to how it was added and by whom.  This data acquisition continues through inspection, repair, test, component replacement, and all the way out to packaging and shipment.

A mountain of context data is gathered.  This is ‘Manufacturing Big Data’.  And all this data is relationally linked, connecting back to the CAD data and the BOM.  The product definition itself, embodied in the CAD and BOM, is the binding element of that massive mountain of data.

When ‘big data’ is considered, traceability required by a given regulatory agency, customer, or market is simply a forward or reverse query against the data set, or a subset of it.

When a manufacturing enterprise adopts a Manufacturing Operations Management (MOM) system capable of managing the entire scope of operations as discussed, traceability is no longer something an enterprise strives to achieve, it is a byproduct.