The time is ripe to radically rethink the application architectures, delivery, and supporting business models from technology providers, consumers, and services firms that connect the dots to deliver complete applications to manufacturing companies to support operations excellence across integrated manufacturing operations. The urgency is to support that growing demand for manufacturing systems that enable demand-driven manufacturing. The global manufacturing community needs to recognize that force-fitting monolithic applications into manufacturing operations is costly, risky, and ultimately creates systems that hinder agility, defeating the original intent. Here’s a quick recap of what’s wrong with traditional MES and ERP monolithic application approaches, or Manufacturing 1.0: - Inflexible MES architectures don’t scale for multisite rollouts.
- Inflexible ERP business processes don’t mesh with the realities of detail manufacturing operations.
- Neither ERP nor MES can cope with constant reconfiguration and requirements to support lean and Six Sigma initiatives.
- ERP and MES have functional inadequacies when faced with multiple manufacturing styles with their own characteristic data models.
- Deployment of MES applications requires significant and increasingly scarce engineering skills.
- Adapting ERP to manufacturing requires domain knowledge that corporate IT business groups don’t have.
- The complexity and cost of MES deployments and traditional hard-wired automation approaches have created an automation threshold, and it’s been difficult to justify the ROI for manufacturing software investments below it.
In short, it’s time for Manufacturing 2.0. AMR Research’s Manufacturing 2.0 defined After several months of refining our ideas on what’s required to bring manufacturing software into the 21st century with manufacturing software vendors, ERP vendors, service-oriented architecture (SOA) technology companies, and our own global manufacturing clients, we now present the details of the Manufacturing 2.0 vision we originally introduced in “Addressing the Manufacturing Backlog: It’s Time for Manufacturing 2.0.” Like Web 2.0, you’ll know Manufacturing 2.0 from its look and feel. Here are a few of its emerging characteristics: - User-centric interfaces that not only streamline activities, but take advantage of available shop-floor talent for deployment, reconfiguration, and software maintenance.
- Manufacturing architectures that capitalize on existing investments by using manufacturing SOA instead of ripping and replacing them with monolithic applications.
- Incorporation of Web 2.0 and Enterprise 2.0 technologies and constructs such as blogs, instant messaging, mashups, search, tagging, wikis, and cool, always-connected mobile devices. This will engage younger generations in manufacturing as well as enhance software usability and collaboration.
- Support for low-cost tags, intelligent sensors, pervasive networks, and mobile workers.
- Event-driven, supply network collaboration (intra- and inter-enterprise) platforms.
- Convergence of product data management and process development models for rapid development of new products and manufacturing processes to accelerate time to market.
We’ve crammed a lot into Figure 1, but it reflects the difficulties that many global manufacturers face. They have to deal with short product lifecycles as well as increasingly complex product configurations produced against short lead times to meet highly variable demand. To further complicate matters, this requires supporting a dynamic network that includes contract manufacturers and outsourced design partners. 
MES applications, most of which are optimized for single-site manufacturing, only support the narrow range of manufacturing styles a particular product was built to support. They take too long to configure, deploy, and manage change. All this exacerbates the manufacturing system’s backlog in the emerging era of demand-driven manufacturing. But hold on—this doesn’t mean MES is dead, not by any means. What it does mean, though, is that new architectures and approaches are needed to orchestrate the narrowly focused and site-specific manufacturing services that traditional MES and other manufacturing operations software applications provide with new supply network composite applications. For a more comprehensive analysis of the new pressures on manufacturing, see “Demand-Driven Manufacturing.” Here’s our concise definition: Manufacturing 2.0 capitalizes on service-based and collaboration-based architectures to let manufacturers dynamically reconfigure sensor and mobile-worker-supported supply networks to make products right first time and on demand. Manufacturing SOA: more need, more real, and different from enterprise SOA...for now Manufacturing 2.0’s supporting technology isn’t in the Web 2.0 or Enterprise 2.0 bags of tricks: it’s manufacturing SOA. While we continue to ask “SOA what?” at the enterprise level (see “SOA What? SOA Seeks a Purpose in an ERP World”), all of the complexities cited earlier have given SOA clear applicability in the heterogeneous manufacturing environment. However, while the promise of manufacturing SOA is real, it largely exists on the Microsoft.NET technology stack, and manufacturing services have evolved in the absence of complete standards and manufacturing master data management (MDM) strategies. If you think MDM and SOA efforts in homogeneous ERP environments are difficult, imagine the complexity of orchestrating services provided by the cast of thousands in the distributed manufacturing applications of any manufacturer’s architecture. 
Central to Manufacturing 2.0 architectures are several key manufacturing SOA components that we first brought together in “SOA on Steroids: The Reality of Manufacturing Composite Applications” and depicted in the AMR Research manufacturing SOA framework (for a comparison with our enterprise SOA framework, see “A Framework Approach to SOA”). While many elements of the manufacturing SOA framework are similar and often identical to enterprise SOA technology components, there are some critical differences. Manufacturing services enablement layer The components in this layer are used in the definition of events and services, the management and orchestration of those services, and the synthesis of information for the purposes of performance visualization and analytics. They include the following: Operations event/activity monitoring Depending on the style of manufacturing, real-time event definition and monitoring may be critical to your composite application. While a failed business transaction can be detected and corrected at a later point, long-term damage to assets, brand, and even loss of life can result if vital manufacturing events aren’t sensed and responded to quickly. Data historians are now a critical foundation for emerging manufacturing SOAs. New-generation data historians do much more than aggregate, compress, archive, and trend real-time data. They are now huge and sometimes distributed state detection machines, designed for complex definition and real-time evaluation of complex events. Operations process management(OPM) Orchestrating manufacturing services isn’t quite as easy as it is for high-level and slow-running business services. As opposed to business process management (BPM), OPM must perform high volumes of complex aggregations and transformations of data from shop-floor sensors and controls, both in real time and predictably. OPM not only needs to link events from sensors and devices but people as well, since they represent the sensors and actuators in many manufacturing steps or intermix with automation. Beyond orchestrating manufacturing services, OPM needs to integrate operational manufacturing processes across disparate time domains to BPM systems, crossing application and enterprise boundaries. This is especially true for quality-driven organizations that can ill afford a massive adverse event that risks their manufacturing assets or, even worse, compliance efforts. Many are not yet convinced that business process execution language (BPEL) can be used to describe and reliably execute high-fidelity manufacturing processes. For more on OPM, see “MES Lite: Building MES Composite Applications With Operations Process Management.” Operations intelligence Just as you need integrated business performance management to see the benefits of BPM, operations intelligence must provide business context to real-time manufacturing processes. This is orchestrated by OPM to provide event-driven, role-based performance management across distributed manufacturing sites (see “Architecting the Next Generation of EMI: Operations Intelligence Meets Business Intelligence”). Operations services bus Like any information bus, this layer consists of the message management, queuing, dispatching, and arbitration functions needed to broker communications between the various entities that share the bus, including legacy applications such as MES and CMMS systems, as well as newer service-based composite applications. The functions to be performed are similar to those of an enterprise services bus. However the enterprise class will struggle to keep up with the volumes and speed required for manufacturing. Manufacturing MDM The system of business processes and technology components that ensure information about business objects (e.g., materials, processes, products, employees, customers, suppliers, and assets) is current, consistent, and accurate whether it is used inside or exchanged outside the enterprise. Schema definition, data model management, and data synchronization across distributed master data repositories are key elements of MDM. Key to manufacturing MDM is the recognition that manufacturing master data has multiple masters within each manufacturing site in addition to coordination points with the enterprise systems that create product definitions. Data synchronization, name space management, and governance processes are necessities that will be supported by emerging operations services bus architectures and federated database management frameworks like Oracle’s Data Hub. Manufacturing composition environment We’re accustomed to the concept of the integrated development environment used to develop run-time execution code for monolithic applications deployed on a single operating system platform. The new world of manufacturing SOA needs a new integrated set of tools that allow us to develop run-time code distributed across multiple platforms, as well as install, invoke, and orchestrate services that are continually evolving on different operating systems and hardware platforms. They must also collect and present data and tasks on various user interfaces and devices. Demand-driven manufacturing forces convergence of the SOA worlds Many manufacturers struggle to force-fit rigid, monolithic applications into the complexity and change that defines today’s demand-driven manufacturing. Leaders have already embarked on innovative approaches to managing manufacturing complexity by using BPM tools, enterprise bus architectures, and federated data management approaches that have their origins with enterprise middleware providers. For most of the innovators, this was a radical departure from the paradigm of standardized enterprise applications deployments. Rather than force-fitting monolithic manufacturing applications, it’s the adaptation of tools and approaches that have traditionally been utilized in high-volume transaction processing in business and finance. Why are they adapting enterprise middleware technologies? Well, put simply, no clear leader or visionary has emerged from the ranks of the automation, MES, quality management, SCADA, or enterprise applications ranks. What ultimately needs to happen is a convergence of the enterprise SOA and manufacturing SOA worlds to allow demand-driven manufacturing to play its pivotal role in demand-driven supply and open innovation networks. We need to build a new generation of supply network applications (see Figure 3). 
Multiple players targeting the Manufacturing 2.0 field The large enterprise platform players and services firms are already jockeying for Manufacturing 2.0 mindshare on the coattails of their applistructure strategies. Let’s look at the players: Oracle Oracle and Microsoft have significant footprints in the manufacturing space at the platform layer and can deliver packaged business processes, databases, and development tools to their customers. Both firms, however, target different sized companies and scenarios. Oracle’s strength is with its increasing set of enterprise and extended supply chain systems. It’s worth noting that virtually all the manufacturing applications offered by Microsoft’s independent software vendors (ISVs) also support Oracle database technology at the operational data store level. Oracle’s prowess in the database and data management arena will work in its favor as companies begin to assemble manufacturing MDM strategies. Microsoft Microsoft’s technologies dominate most manufacturing scenarios either through core technology offerings (like BizTalk server, SQL databases, and the SharePoint portal), Office applications (like Excel and Access), or via partners such as Apriso and Invensys. That said, it remains to be seen whether Microsoft can overcome its “platform for the masses” market approach long enough to offer the database, model, and event management services needed to cement its role as the prominent Manufacturing 2.0 platform provider. We’re still surprised that the company hasn’t already stepped up to deliver the Manufacturing Composition Environment in AMR Research’s Manufacturing SOA framework. With many manufacturers believing that MES is the acronym for Microsoft Excel Spreadsheet, this is Microsoft’s opportunity to lose. SAP SAP has the core ERP functionality sewn up for many large global manufacturing organizations, but has yet to prove its applistructure mettle. Rather than develop the core manufacturing functionality and services itself, SAP has developed an expansive partner ecosystem, relying on them to deliver composite applications though SAP xMII. As we hinted in “SAP To Resell Visiprise MES, But at What Cost to Manufacturing 2.0?” SAP is well positioned to capitalize on its impressive manufacturing partner network and deliver on the promise of Manufacturing 2.0. We highlighted the former Lighthammer as one of the pioneers of OPM and perhaps a Manufacturing 1.5 capability (see “Composite Applications Find a Home in Supply Network Operations”). With SAP’s commitment to continued investment and exploiting synergies with NetWeaver, as well as a “Duet for Manufacturing” agreement with Microsoft in the mix, xMII could be transformed into a more comprehensive manufacturing composition environment for Manufacturing 2.0. IBM Big Blue is an intriguing unknown in this equation. IBM’s long history as a middleware provider coupled with its strength in BPM (not to mention its manufacturing architecture experience garnered through the global services group in the chemical process industries) combines to create a compelling scenario that goes something like this: IBM comes in and builds your manufacturing infrastructure, constructs the bus, creates the MDM layer, and then hosts your core enterprise applications. Coupled with the MRO Software acquisition, it also offers to guarantee uptime for a fee. This could be a reach, perhaps, but one that has merit. IBM has also partnered with Cisco and Rockwell Automation. It uses WebSphere to integrate Rockwell programmable logic controllers (PLCs), motion controllers, and other automation components with FactoryTalk Production Center. IBM also uses Cisco to best facilitate data capture, visibility, and analysis across a network of multiple manufacturing assets to drive greater flexibility and productivity. For example, in the automotive industry, this can potentially lead to higher quality levels through better process interlocking, error proofing, and process interlocking for early warning system capabilities. Integration of this data with enterprise systems can lead to better traceability of supplier and OEM data to slash annual warranty exposures. The role of services firms Services firms like IBM Global Services, Tata Consultancy Services (TCS), Satyam, and Wipro will have to rely on Oracle and SAP customers to buy into Manufacturing 2.0 in order to bring their manufacturing and SOA expertise to the table. TCS has closely aligned with SAP’s adaptive manufacturing vision and has already created the first certified composite application on SAP xMII. TCS’s RapidSigma is aimed at simplifying the user experience for Six Sigma data analysis on the shop floor by helping process engineers to see cause-and-effect relationships of critical parameters of the production process and identifying the factors that can change production. This helps improve the to-the-middle lifecycle of continuous improvements with historical trends. TCS has also created a portfolio of add-on components on xMII through work with other clients. It includes a composite for asset-intensive environments that integrate TCS’s EntICE framework with multiple pieces of equipment and DCSs via SAP xMII for monitoring equipment health and other thresholds. It also uses web services to connect with SAP’s Plant Maintenance module. Manufacturing 1.0 incumbents, embrace Manufacturing 2.0 pioneers While we and the market wait to see how the large enterprise players will jockey to own the Manufacturing 2.0 opportunity, this doesn’t mean the opportunity for traditional manufacturing vendors has been eclipsed. If anything, they will still be called on to provide the underlying services we have outlined. The following vendors will continue to benefit from the rediscovery of manufacturing: - Automation vendors—Citect, GE Fanuc, Invensys, and Rockwell Automation
- Pure-play MES providers—Aegis Industrial Software, Camstar, Eyelit, iTAC, Werum, and Visiprise
- Specialists—InfinityQS in the quality market and Pavilion and AspenTech in the advanced process control and simulations market
If anything, manufacturing vendors looking to play a role in the industry transformation or potential consolidation that is underway should continue to embrace the potential market leadership the enterprise players possess with Manufacturing 2.0. If you have questions about AMR Research’s Manufacturing 2.0 vision, let us know at sno@amrresearch.com.
| 
;)
;)
;)