Many manufacturers attempt Continuous Improvement (CI) initiatives without an aligned MOM system strategy and fail; only an 8% average improvement according to the London School of Economics (LSE). As well, many manufacturers attempt MOM systems without an aligned CI initiative and fail; only a 2% average improvement according to LSE. The same LSE Study found that that the combination of CI and MOM resulted in a 20% average improvement.

As a result, a new global industrial revolution has begun optimizing manufacturing plants by combining CI methods with an adaptive MOM system architecture that are put into place using a Manufacturing Transformation Strategy.

Successful optimization centers on prevention of production issues in the MOM design by characterizing operations work processes through Lean and Six Sigma methods. Based on the resulting user and system functional requirements, a MOM System Roadmap is derived including risks and a tangible return on investment (ROI) identified.

However, the MOM System Roadmap must include cultural transformation requirements for each system, to cement the adoption by all stakeholders or risk project failure.

Non-value-added operations costs are caused by poor MOM system planning. Isolated MOM systems, while solving specific manufacturing issues, ultimately result in high operations costs due a fragmented, constrained architecture and confused MOM master data that can’t be shared across the enterprise. These additional passed-on costs from inadequate system design and governance range between 100 to 300 times more expensive than determining requirements and architecture through a structured manufacturing transformation assessment.

There is a simple answer but complex solution. Plant scheduling must be done on maximizing profit per order and not just on line balancing. While knowing the margin and customer-priority for all WIP orders in the daily mix, the high margin orders are prioritized to maximize profit per day.  Detailed scheduling needs to focus on margin and customer priority. This requires the manufacturer to understand that even the best lean-pull plant in the world must still be able to respond quickly to adverse negative events to make money.

During every shift, this perfect manufacturer still has quality events, downtime events, raw material issues, and uncharacterized processes requiring change in process and design. However, the best-in-class have designed their MES and plant floor systems to provide real-time feedback to their schedulers/controllers on all adverse events. When the scheduler/controller also has real-time feedback on the response time to reconcile each non-conformance or resource issue, top-down feedback on event and the time-sensitive response are applied by shop floor personnel to rapidly reconfigure the plant floor.

Dynamic scheduling requires that pre-engineered alternative routings are scheduled and deployed to line supervisors.  This is where the best-in-class manufacturers separate from the other 85%. Their supervisors have visibility into the state of orders, resources, order margin, and order priority on shift and are trained to dynamically respond to a new schedule release in less than 30 minutes with the resource deployments to the new line configurations.

The supervisors and operators as knowledge workers know that they must fill orders with the highest margin based on what WIP is in play. For example, if the shift is started with 4 lines with 5 orders per line to process and then lose a line and loss 3 orders due to poor quality, profit margin for that shift is directly proportional to the speed of reconfiguration. This requires finite capacity scheduling with plant floor systems providing accurate WIP and resource feedbacks.

Heijunka and other non-system lean methods are often brought up as the best detailed scheduling approach. These lean method without systems simply cannot balance or level a line if the order mix has a high number uncharacterized MTO or ETO work processes and if the SKU count is high (ex 50 FG SKUs).  Heijunka and other pull methods work well in characterized MTS, medium to low SKU count situation.  But even though lean methods are designed to respond to bottlenecks events, the methods simply can do not consider the activity-based-cost and proportional costs that must be part the Dynamic scheduling requirements to develop and algorithm to determine the best line reconfiguration based on profit margin.

Manufacturing Interoperability is the ability for MFG systems to enable organizations to work together (inter-operate). MFG Interoperability is a property of a system architecture where interfaces are completely understood to work with other enterprise systems, present or future, without any restricted access or implementation.

Several criteria can be used to discriminate between systems that are “really” inter-operable and systems that are sold as such but are not because there is non-disclosure of one or several interfaces implementation or access restriction built in the product/system/service.

Syntactic interoperability: Syntactical interoperability is a necessary condition for further interoperability. If two or more systems are capable of communicating AND exchanging data, they are exhibiting syntactic interoperability. Specified data formats, communication protocols and the like are fundamental. XML or SQL standards are among the tools of syntactic interoperability. This is also true for lower-level data formats, such as ensuring alphabetical characters are stored in ASCII format in all the communicating systems.

Semantic interoperability: Beyond the ability of two or more computer systems to exchange information, semantic interoperability is the ability to automatically interpret the information exchanged meaningfully and accurately in order to produce useful results as defined by the end users of both systems. To achieve semantic interoperability, both sides must refer to a common information exchange reference model. The content of the information exchange requests are unambiguously defined: What is sent is the same as what is understood.

Interoperability and Open Standards: Interoperability must be distinguished from Open Standards. Although the goal of each is to provide effective and efficient exchange between computer systems, the mechanism for accomplishing that goal is very different. Open Standards imply interoperability by definition, while interoperability does not. Open Standards imply a wider exchange between a range of products or similar products from several different vendors or even past and future revisions of the same product. Interoperability may be developed post-facto, as a special measure between two products, while excluding the rest, or even when the vendors of the two ‘interoperable’ products are forced into a dominant/submissive or exploitative relationship vis-a-vis themselves or their customers.

Open Standards rely on a broadly consultative and inclusive group including representatives from vendors, academicians and others holding a stake in the development. That discusses and debates the technical and economic merits, demerits and feasibility of a proposed common protocol. After the doubts and reservations of all members are addressed, the resulting common document is endorsed as a common standard. This document is subsequently released to the public, and henceforth becomes an open standard. It is usually published and is available freely or at a nominal cost to any and all comers, with no further encumbrances. Various vendors and individuals (even those who were not part of the original group) can use the standards document to make products that implement the common protocol defined in the standard, and are thus interoperable by design, with no specific liability or advantage for any customer for choosing one product over another on the basis of standardized features. The vendors’ products compete on the quality of their implementation, user interface, ease of use, performance, price, and a host of other factors, while keeping the customers’ data intact and transferable even if he chooses to switch to another competing product for business reasons.

There is a number of manufacturing operations integration standards that have been developed over last 20 years and are ready to use!  But in determining the standards to apply to a given manufacturer’s plant, typically a set of integration standards have to be chosen and applied as a work set to establish the basis for MFG Interoperability.

3 Types of MFG standards required for MFG Interoperability:

1. Business Document Format standards to define the data definitions, process definitions, and data structure: Examples: XML/XSLT, OAGIS, EDI, B2MML
2. Messaging Protocol standards for providing basic messaging framework: Examples: OPC, SOAP, JMS, AS1, AS2, RNIF, FDT, EDDL
3. Message Transport standards for communication protocol: Examples: HTTP, HTTPS, FTP, SMTP, TCP, UDP, MQ, MQTT

Many manufacturers come at this challenge from only top-down business intelligence methods with financial metrics based on monthly mass balance numbers and/or basic macro supply chain metrics like capable-to-promise. This approach is driving by looking in the rear view mirror. It does not factor in the current as-is plant capability, capacity, and cost metrics into a high value decision making process. Global supply chain management requires near-real-time plant metrics to be able to understand many dependent business and operations processes. Based on the current state of operations resources and demand, management MUST know how best to balance the flow and purchase of materials from suppliers to customers to ensure the highest profit margins on every sales order.

One of the best practices for doing this alignment is to build the top-down and bottom-up metrics frameworks in parallel. Let’s start with the top-down best practice first. The Supply Chain Council’s Supply Chain Operations Reference (SCOR) Model has spent over 10 years developing their Metrics methods and mapping. SCOR PLAN Level 1 Metrics map to supply chain operations Level 2 metrics which then for the plant maps to MAKE manufacturing Level 3 metrics. The PLAN Level 1 metrics are chosen based on the focus type that your company’s supply chain must operate. PLAN Level 1 metrics categories are reliability, responsiveness, flexibility, cost and assets. So if your supply chain focus is on reliability, you would construct the Perfect Order Level 1 metric which maps to four Level 2 supply chain process metrics and which in turn are partially supported by five MAKE Level 3 plant metrics. A note here….the Levels in SCOR and not the same as the ISA-95 Levels but SCOR Level 3 is the same as MOM Level 3 in ISA-95.

So now that we have the SCOR Level 3 plant metrics, we do the bottom-up performance metrics for the plants manufacturing operations to determine what metrics are to be used to construct the SCOR metrics. Big note here…..the SCOR Level 3 calculation is not the same across all plants or lines because each manufacturing form or type uses different performance metrics to control the operations. An example of this is the throughput for a discrete packaging line is calculated in units per hour where throughput for a batch mixing operation is calculated in pounds or gallons per mix per hour.

Now, to do the bottom-up metrics framework for a given plant, the best practice is to start with a value stream mapping (VSM) of the end-to-end process and determine the waste streams to be controlled, their losses per use cases, their performance metrics, and the cost per waste stream. Second, around each work center’s or unit’s actual physical process, Six Sigma analysis must determine the process variances, their losses per use cases, their performance metrics, and the cost per event type. The combination of VSM and Six Sigma to determine the bottom-up metrics framework provides the requirements definition for the MES that must do the data collections, calculations, reporting, and interfaces to construct the SCOR Level 3 metrics.

Hope this helps.

Analytics and reporting is another area greatly dependent on an organization’s master data. Master data is used by several functional groups and stored in different data systems of record across an organization and may or may not be referenced centrally; therefore, a high possibility exists for duplicate and/or inaccurate master data.

Thus Master Data is persistent, non-transactional data that defines a business entity for which there should be an agreed upon view across the organization.

The Big Issue is that Enterprise Master Data and plant-side manufacturing master data are not aligned. This creates wide spread data integrity issues for business and manufacturing intelligence. For example, the same data element (equipment name), but not common data object (multiple name objects), is present in different systems and at different levels of the systems hierarchy. This data is often different and not standardized between enterprise and plant’s department-based systems.

MES/MOM (Level 3) and SCADA/process control (Level 2) lower-level systems hold more detailed master data information than the ERP, SCM, and SOP systems. An enterprise MDM system is thus unsuited to keep and maintain detailed manufacturing master data. The table below shows different master data types from business process to operations processes to work processes on the shop floor. As many manufacturing company evolve the manufacturing intelligence, they are discovering the data integrity require highly prescriptive governance to synchronize master data by objects, artifact version, and type to maintain data integrity for the detailed scheduling, production order genealogy, and performance reporting.

Table: Mfg Master Data Examples
S&OP PROCESSES
Enterprise MD: Resource Planning & Allocation
MASTER PRODUCTION SCHEDULING & RESOURCE PLANNING
Enterprise MD: Master Schedule, Work Orders, BOM, General Recipe
SITE SPECIFIC RESOURCE PLANNING & SCHEDULING
Mfg MD: Bill of Materials, Bill of Process, Bill of Equipment, Bill of Assay/Test, Labor & Skills, Bill of Compliance
SITE SPECIFIC ROUTINGS, WORK DEFINITION/DISPATCH
Mfg MD: Master Recipe, Formula Optimization, Scaling & Substitutions, Detailed Work Instructions, Detailed Production Schedule, local labor Laws & Regulations
SITE SPECIFIC EXECUTION CAPABILITY (EQUIPMENT & ASSETS)
Mfg MD: Control Recipes, Equipment settings, Maintenance Requirements, Tooling Management, Equipment Operating Specifications, Calibration, Process Parameters, SOPs, Equipment State Models for each OEM, and Phase Sequencing and Logic

While 20th century manufacturing was about minimizing labor and materials used per produced unit, 21st century manufacturing is about minimizing water, energy and waste. Many think tanks view the 21st Century as a water economy, not an oil economy. Consequently, a manufacturing plant MUST become highly intelligent with some level of artificial intelligence.

When the “normal” work processes and product quality begin to or become abnormal, the operations management systems controlling the real-time work process and material transformations are now being designed to address the corrective paths based on available real-time resources, alternative routings, rework routings, waste stream routing, and the cost of each option.

The physical materials transformation processes and facility resources must coordinate by the Manufacturing Operations Management (MOM) systems. This means that MES does more than expanded MRP+ where a production order is simply scheduled to assigned resources (equipment, materials, and people) and then the product routes and recipes are executed and validate by step and BOM per each operation. MES becomes MOM where production processes are actively coordinated maintenance operations, quality operations, and inventory movement operations.

Comment from Bob Corker: The concept of an intelligent plant is really brought into focus when one realizes that the cost of resources such as energy, water, and transportation as well as materials cost will be (and are now) highly variable. So as the resource and materials cost vary on a daily and even hourly basis, the cost of manufacturing must be a primary variable in production scheduling.

Comment by Carl Green: The results are already being reported in water minimization by companies such as Nestle and Coca Cola who have reduced their water use by 30-50% for the same levels of production at the same cost. I do not have the references for this but it is out there. So “sustainability” is not just a buzz word but a strategic and tactical initiative.

MOM systems are the pivot around which other ERP, PLM, SCM systems revolve. With a MOM architecture to control and optimize plant work processes and provide real-time, high integrity data into the ERP, PLM and SCM algorithms. These enterprise systems and the business processes that they are supporting are simply making high dollar decision based on dated or incorrect data on the status of orders, plant capabilities and capacity. Without MOM systems, these other applications within the manufacturing landscape operate blind or, at best, have only outdated information to work with.

This importance of MOM systems within the manufacturing IT landscape indicates the complexity of the issues and the processes that need to be coordinated from Localized KPA/KPIs. Collaborative manufacturing and the related improvement projects lead to localized optimums that can conflict with company strategy and goals or the goals of other processes in the value chain. Collaborative Manufacturing must drive to characterize and then reduce conflicts between departments or processes.

Comment from Charlie Gifford:
There are three domains to continually align when looking at a manufacturing company:
1. Order fulfillment on the demand side
2. Material Movement on the supply side
3. Scheduling and execution of plant work processes.

The systems and their supported business processes that live in each domain need a collaborative infrastructure to tie the systems together. People have always looked at the systems for these domains using a two-dimensional model. The two-dimensional model fails to depict the complex relationships and dependencies between applications and the business processes and work processes that they are supporting within a manufacturing company. Consequently, the 3 domains actually inter-operate across a 3D model where MES/MOM systems actually support the business processes of 1D) PLM from design to customer support, 2D) ERP from sales to production, and 3D) SCM across the 3 domains.

In order fulfillment domain:
1) the demand-side requirement for flexibility conflicts with the mfg-side need for process stability and minimum product change-over.
2) In addition, the demand-side requirement of adequate stock levels conflict with the mfg-side need for minimum stock levels and maximum stock turns.

In material movement domain:
1) the supply-side need for the suppliers and 3PL providers to have long lead-times conflict with mfg-side Purchasing’s need for maximum flexibility.
2) The supply-side need for raw materials warehouse to carry minimum stock and maximum stock turns conflict with mfg’s need to have maximum stock so as to not run out of raw materials.

In the Detailed Production Scheduling and Shop Floor Work Processes domain, the demand-side requirement for schedule flexibility conflicts with the mfg-side need for process stability and minimum product change-over. The mfg requirement of adequate raw materials stock levels conflict with the demand-side need for minimum stock levels and maximum stock turns. The need for Final Product warehouse to carry minimum stock and maximum stock turns conflict with sales’ need to have maximum Final Product stock so as to not run out.

Comment/Reply from Alan Foster:
Collaborative Manufacturing is the supply chain “process control” where intelligence is collected, analytics are applied, and strategic decisions are based on high data integrity. The 3D business processes within the company need to be optimized to reduce the cycle time for inventory-flow, information-flow and cash-to-cash flow. The faster these 3D business processes flow, the more lucrative and cash-rich the business will be. If information is not available in real-time, it will slow down both the inventory- and cash-flow processes. If the information is inaccurate (poor data integrity), it will reduce the reliability of the processes and slow down the flows. If the final invoice is inaccurate, even if the products have been delivered on time as per schedule, the cash-to-cash cycle will not be reduced as queries will be made, credits approved and passed and only then will payment occur. Speed and accuracy is thus critical to Collaborative Manufacturing.

Global MES implementation teams must understand that each plant has its own manufacturing form of work processes and personnel culture based on “What it Makes, How it Makes It, Where it Makes it, and Who it Makes it for”.  Understand the plant environment has its own product set and order mix, SKU count, the customer requirements, product lifecycle, the supply chain(s) that it operates, and HOW it makes products.  The global implementation of MES is, of course, straight forward if each plant makes the same products in the same supply network for the same customer. But most plant fleets are very diverse from material processing, to parts fabrication, to sub assembly and final assembly, to testing, to packaging, and to finally MRO (maintenance, repair, and overhaul). This is not to say at all that an MES cannot be designed to globally scale across multiple plants. It is to say that the MES functionality set and work processes are different from plant to plant and line to line.  If you model the real-time work processes and supporting information flows for each operation across 2 plants, you will find only 20-40% common work process and data exchanges. But if you model 3 plants, the common work processes and the shared MES functionality set to support them go up to 50-80% between any two plants.

Many innovative manufacturers over the last 15 years that have done global MES implementations have found that when 4 to 5 plants are modeled simultaneously that common reusable work processes and shared MES functionality goes to about 85% where each reusable work process is used in at least 2 plants.

However, the biggest issue for global MES implementation is not just matching reusable MES components to each plant’s manufacturing form; it is the transformation of the each plant’s culture and maturity in the areas of continuous improvement and integrated work process systems.  You can bring the horse to water…. But are the plant personnel, work processes, and their current organization structure ready to support the new MES functionality that you have globally deployed.

Most Global MES implementations do not get past the first 3 plants because the team has not figured out how to transform the current culture to one that adopts and uses the deployed MES CORRECTLY and Effectively.

Comment from Rod Kelly: Can single set of financial metrics be applied across a global set of plants?  You need to have at least a financial basis of comparison across your plants and business.

Comment from Alan Foster:  Yes, one of the first steps in doing a global MES implementation is to design a Mfg Metrics Framework for both financial and operations standard metrics. This standard set of metrics must come from the Mfg Transformation Assessment of the each form of mfg operations and supporting work processes so we understand the KPIs at the operations (throughput, quality, efficiency, etc.) and financial (cost per unit, labor, materials, facility, energy, etc.) areas.

Comment from Mike Gray: How does a Global MES implementation consider the current pains and needs at each plant and their specific continuous improvement plant? Many global MES implementation that I have seen are partially rejected by each plant because the globally MES application does not have functionality that is addressing their specific set of operations pains and needs.  So each plant develops their own “shadow” IT in custom Excel applications, paper forms, or database applications to help the plant characterize and fix the current problem set.