Curbing MESsy Shop Floor State of Affairs - Part I




Those that follow manufacturing-oriented enterprise applications have likely noticed for some time an uptick of conversations about the need to better integrate high-speed manufacturing operations (the real-time world of the plant) with the planning and engineering departments (the transactional and design world of enterprise systems). The nirvana (or utopia) hoped for thus far has been to provide a single point of operation and control for manufacturers to: Plan, Define, Control, Execute, and Analyze Production.

Why do we need integrated manufacturing operations, or manufacturing execution systems (MES) linked to transactional enterprise resource planning (ERP) systems, likely via some plant-level integration hub and visualization & intelligence layer?

Well, it is not a major revelation to say that, for instance, in the discrete manufacturing sector, fabrication and assembly processes are being run and managed by isolated applications, such as “Post-it” notes, Microsoft Excel spreadsheet, Microsoft Access databases, and a plethora of niche vendor’s plant applications (point solutions like data historians). This creates an overwhelming number of individual silos (or islands) of manufacturing data and operations.

These silos are typically not connected to enterprise-level (“ivory tower”) systems like ERP, Computer-Aided Design (CAD), Product Lifecycle Management (PLM), and so on. This lack of integration and real-time connection then all too often results in huge operational inefficiencies, lost productivity, wasted time and materials, sub-par products, and so on.

Consequently, major decisions in the offices are based on theory and hunch rather than on actual and actionable data. But instead of traditionally managing operations “in the dark”, companies should rather strive to capitalize on all of the operational opportunities coming from the following sources: people, processes, and the plant equipment.

ERP Does It... Not!

Some might logically wonder whether ERP systems can take care of this (and why not, if that is the case). Well, at best, the core ERP systems’ functional scope only provides a financial and inventory snapshot of how a manufacturer is performing. Core ERP systems cannot tell users what is happening on the manufacturing floor right now (at this instant). Ironically however, what is happening at this moment impacts the financial performance later.

To be more illustrative, ERP is good at producing a forecasted demand plan by decision makers, and giving answers to sales, purchasing, and manufacturing orders’ inquiries like “What?”, “When?”, “For whom?”, and “At what cost?” Conversely, MES is good at providing the record of production that is supplied by plant operators (e.g., engineers, supervisors, machinists, etc.), who thereby inadvertently turn into mere data collectors.

The execution system is able to provide answers to the questions like “What are the schedule changes?”, “What is the product build history?”, “When will it be done?”, “How is the product quality?”, and “Where is the batch?”, but without any awareness of the customer (the particular order for that customer) or the particular order costs. In a nutshell, MES systems are devoid of any customer- and order-related information.

By inserting an informed schedule and manufacturing intelligence layer between the enterprise and the plant, enterprise planners can continue to make strategic decision-making, while the operations staff can transform from mere data collectors into tactical decision makers and action takers. But, that is not simple, as discussed in TEC's earlier article entitled “The Challenges of Integrating Enterprise Resource Planning and Manufacturing Execution Systems.”

Let me try to be a bit more detailed and depict some necessary data and process flows between ERP and MES systems (and vice versa). Well, the production order release process would logically go from ERP to MES, with a raft of order master data to be passed, such as: Order Information, Serial Number, Material Master, Labor Charge Code, Routing Operation, and Bill of Material (BOM).

To be even more specific, in case of the high-tech/electronics sector, additional order information would include: Traceable Parts, Assembly Points, PRT Work Instructions (i.e. for wiring), and PRT Attachments. Any subsequent order changes in terms of quantity, order placing and delivery dates, splitting orders, engineering change number (ECN) cut-in dates, etc. would also pass from ERP into MES.

Conversely, MES would provide the order feedback for ERP in terms of Completions, Scrap, and Reporting individual operations. Getting even more granular, MES could provide the individual operation feedback for ERP in terms of: Completion, Scrap, Operation Start Times, Reporting Operations, Rework Quantity Reporting, First-time Completion or Operation Completion Reversal. Last but not least, detailed scrap data such as at the Unit Level or Component Level, or Reuse of scrap (i.e., “un-scrap”) would also go from MES to ERP.

Hooking ERP Up with MES: Good, But Not Sufficient Yet

Without such a tight and near real-time integration, there is much anxiety and frustration within any enterprise that is in search of a more competitive, profitable, safe, and agile factory. How can any manufacturing company reduce non value-adding administration and empower their workforce to take immediate remedial actions?

Namely, the typical current state of affairs from the perspective of a senior vice president (SVP) of operations could be summarized as follows:

  1. On one hand, the ever more pressured manufacturing environment demands acceleration of the stock-keeping units (SKU) mix and shorter lead times, all due to ever more demanding and fickle customers; but

  2. On the other hand, the real world situation is of little overall enterprise and/or SKU-level profit visibility, and the company has to rely on (suboptimal) average key performance indicators (KPIs), with emergency scheduling (constant firefighting) on paper or Excel documents.


In such “clueless” environments, there are “blind spots” everywhere in terms of determining yields and losses, hidden capacity opportunity, and masked process routing and constraints by reactive work practices. Also, there are increased risks of quality non-compliance leading to manual quality assurance (QA) processes, whereas continuous improvement efforts are floundering and remain unmeasured. In a nutshell, the hands-on plant people do not seem involved and are ironically not accountable for what they should be.

The future state should logically be the inverse of the above, and the usual “first remedial step conclusion" is to gather the glut of data from data historians and MES databases, and then decide what to do. But, without smart and intuitive plant applications that have visualization and contextual business intelligence (BI) capabilities (and that are thus accepted by the plant staff), this will all be yet another exercise in futility.

The reality check reveals an “inconvenient truth” that many MES investments fail to deliver hoped for performance management outcomes due to people issues. Namely, after 18 months or so, the embattled company in case might have an overall equipment effectiveness (OEE) dashboard that the plant engineers occasionally look at (and which might have cool colors on it), but without a pervasive effect (actionable info) and acceptance across the plant (and entire enterprise).

Part II of this blog post will analyze some possible remedial approaches, and provide examples of ERP vendors with good native MES capabilities too. Your views, comments, opinions, etc. about any above-mentioned solution and abut the software category per se are welcome in the meantime.

We would also be interested in hearing about your experiences with these software solutions (if you are an existing user) or your general interest to evaluate these solutions as prospective customers.
 
comments powered by Disqus