MES: From Device to Decision
Manufacturing management increasingly is looking upon the plant floor as a source of untapped wealth. Value there lies in the rich vein of data streaming through automation control systems—data to be mined to improve critical enterprise performance. That means boosting global operations and supply chain efficiency; coordinating collaboration among partners and suppliers; increasing market...
Manufacturing management increasingly is looking upon the plant floor as a source of untapped wealth. Value there lies in the rich vein of data streaming through automation control systems—data to be mined to improve critical enterprise performance. That means boosting global operations and supply chain efficiency; coordinating collaboration among partners and suppliers; increasing market responsiveness and profitability; and complying (cost effectively) with regulations.
Both sides—corporate management and engineering—agree on the benefits. The issue is how best to efficiently extract plant floor data for maximum value. That's why engineers increasingly find themselves as much at the center of automation-to-enterprise integration discussions as the IT department.
After generating a lot of interest in the late 1990s, manufacturing execution systems (MES) are again being viewed as an attractive means for achieving a viable device-to-enterprise information flow. According to Boston-based AMR Research, the MES market broke $1 billion last year for the first time in its 20-year history. A convergence of major market drivers and a blizzard of regulatory mandates (Sarbanes-Oxley, WEEE in electronics, TREAD in automotive, the Bioterrorism Act for food, and 21 CFR Part 11 in pharmaceuticals) have created what AMR likens to a 'perfect storm' for MES demand.
Of paramount value is the unparalleled visibility MES provides for exploiting plant floor data not native to higher-level systems. 'For a long time, business systems weren't geared to use information out of production systems,' says Matt Bauer, director of MES for Rockwell Automation. 'Now they're hungry for it, able to make extraordinary use of it. Management wants more visibility to improve global enterprise performance.' Bauer is known for referring to MES' funneling of data to higher-level business systems 'feeding the beast.'
Data context granularity
While some automation equipment status data can have direct value to ERP (such as denoting whether a tank is empty and available), and it's possible to port automation data directly to the enterprise layer, typically via data historian software, MES provides a more crucial intermediary function between control and enterprise systems, especially in complex production environments.
'I've heard of people doing direct ERP-controls integration, but most manufacturers are afraid of that option,' says Francois Leclerc, business manager, Business.Flex PKS solutions for Honeywell. 'The issue is cyber security. And if they are doing it directly, they're having to create some sort of MES function themselves that scrubs and transforms the data to be useful to ERP.'
Reasons behind manufacturers’ MES purchases in 2005.
This intermediate step of transforming production data is critical because 'business systems aren't designed to handle every event and alarm from automation equipment,' says Claus Madsen Abildgren, production management solution manager for Wonderware. 'MES provides a contextual format for the data. It aggregates and provides higher-level granularity of real-time data, posing data in a form more easily understood by business systems.'
MES typically offers modular functionality for resource management and allocation, order dispatching, data collection/acquisition, tracking and traceability, quality, maintenance management, performance analysis, and scheduling. It gathers data from the control layer (such as equipment utilization and material consumption) and provides contextual meaning for ERP systems, using ERP data (such as bills-of-material and weekly schedules) to prioritize and direct daily shift work.
A compelling attribute of MES for many manufacturers, says Julie Fraser, principal with Industry Directions, a market research firm in Newburyport, MA, is its ability to 'see interdependencies among production lines. Manufacturers can optimize each area and still have a sub-optimal whole. Addressing the interdependencies among processes is where MES has really filled the gap,' she explains.
Long considered a tool primarily for use in manufacturing tracking and tracing, MES now finds itself at the center of the extended manufacturing universe because of the production insight access it offers.
Tyson pilots MES
At Springdale, AR-based Tyson Foods, however, it took only a one-line pilot in one plant to convince hesitant executive management to 'green light' a multi-plant MES rollout. Chad Couch, vice president of manufacturing systems, had requested $2 million to apply MES across an entire poultry-processing plant pilot, but was granted only $100,000 to make an MES business case on one line.
Couch's goal was to align all business and production processes to give the customer the highest quality product at the lowest price. To do that, the company needed faster access to production information for decisions affecting efficiency, quality, and cost. Tyson selected Siemens' Simatic IT MES solution, in part because of the tight integration it provided with SAP, Tyson's corporate ERP system.
Process velocity—the time to get information to a decision maker—sped up dramatically using Simatic IT linked with SAP, says Couch. 'We did a complete flip-flop from moving information very slowly to very quickly.'
According to Couch:
Real-time information access for analysis jumped from 0 to 37.7%;
Information by minute climbed from 3.7% to 15.8%;
Use of hourly information declined from 56.7% to 39.3%;
Use of daily information dropped from 37.7% to 6.9%; and
Use of weekly 'bucketed' information decreased from 2% to 0.2%.
A chief attraction of the Siemens/SAP combination for Tyson is that both embrace the ISA 95 standard. 'We have one ERP system and one MES. And thousands of PLCs and SCADA systems,' says Couch. 'Follow the ISA 95 standard. It'll save you a lot of work.'
Integration standards greatly facilitate inoperability between disparate systems from different vendors by separating process application functions from transport and communication functions, and by providing models, terms, and definitions. These standards—such as industrial networks to link automation devices; OPC for linking devices, SCADA, and DCS; and ISA 88 for batch recipe control—effectively shift more of the integration burden from users to vendors and have greatly accelerated new technology adoption in the industry.
'Interoperability among the control layer and MES and ERP has been a long-time concern for customers,' says Karsten Newbury, general manager of Siemens' Process Automation and MES Solutions. 'ISA 95 is becoming a force for interoperability between ERP and the shop floor like ISA 88 has been in the batch world.'
ISA 95 is fast-tracked to hasten integration between MES and ERP, patterned on the long-standing success of ISA 88. While ISA 88 took years to be widely adopted, acceptance of ISA 95 is rapidly gathering momentum with endorsement by IT vendors, such as SAP and Microsoft. (See 'ISA 95 Model' sidebar.)
Most leading automation vendors endorse ISA 95, particularly those moving assertively into the MES space through acquisitions of best-of-breed niche vendors. Automation vendors with MES offerings are well positioned in the market against ERP vendors because of their deep roots in real-time data management and long-standing relationships with plant operations management.
'ERP vendors have something of a credibility gap. When they talk about real time, they mean sometime today,' says Chris Boothroyd, Honeywell product manager. 'We have the heritage that allows us to design applications and support decisions closer to real time.'
Some automation vendors are leveraging MES acquisitions by developing overarching, standards-based platform infrastructures that provide common services like workflow, security, and alarming. Siemens acquired a foundation in this area with its acquisition of ORSI, which had designed its core product around ISA 95. Siemens has since invested $250 million building up its Simatic IT framework that provides for common plant modeling, workflow configuring, and other services. Likewise, GE Fanuc and Rockwell are also augmenting acquisitions by investing in platform architectures to develop their Proficy and Integrated Architecture offerings, respectively.
As for internal developments, Honeywell is in the process of building out its Experion platform, ABB similarly continues to extend its 800xA platform, and Wonderware has positioned Invensys's ArchestrA, as a multi-plant connectivity infrastructure platform.
Arla Foods, the largest dairy company in Europe, is using Wonderware's Plant Intelligence MES solution at its Christiansfeld Dairy Centre in Denmark to link all its automation devices for greater control and traceability. ArchestrA has been implemented there with MES to link all of the company's dairies into one global multi-plant intelligence system. This technological adaptation has been a key strategic move for Arla, which has grown significantly through acquisition, now supporting 70 European processing plants.
The ArchestrA-enabled 'Arla One' system enables them to consistently measure online equipment effectiveness and benchmark progress of all 70 sites. As a result, managers can make strategic decisions based on what each plant can accomplish.
'Previous dairy systems were not even automated,' says Erik Veslov, production manager of the Christiansfeld plant. 'PLCs were not connected, there was no reporting and no traceability. We can now provide critical process information to people who need it to make important decisions. In addition to faster response times, this new level of traceability actually reduces the risk of having to recall a large amount of product because we have detailed information on each batch, enabling us to report efficiency, quality, and performance data and key performance indicators in a uniform way.'
Analysis of best plant practices shows that those using MES achieved faster productivity improvement and larger yield gains, while reducing costs more dramatically, according to Fraser of Industry Directions. It [MES] also contributed to fourfold greater profitability, she says.
Part of these enormous gains can be attributed to the fact that 'those who use MES also use other technology and best practices as well,' Fraser says, 'but part of being a leading-edge manufacturer is having a plant-wide information system—being able to aggregate information that comes out of control and automation systems and make it more digestible to IT systems.'
Research by AMR builds on Fraser's assertions. The analyst firm reports that MES has been shown to pay for itself on traditional cost reduction measures within 6 to 24 months of going live. Returns accrued over the long-term, however, far exceeded initial justification hurdles, with some manufacturers seeing returns in the 6x to 10x range per year.
As impressive as MES is, it is important to remember that it is the rich vein of data in the automation layer where the raw wealth lies. MES is simply a means of efficiently extracting and presenting it to engineering and management decision-makers who can parlay its value.
Frank O Smith is a freelance writer who has covered manufacturing systems integration for more than a decade.
ISA 95 model explained
ISA 95 is an evolving reference model defining good integration practices for linking enterprise and control layer activities based on a four-level manufacturing enterprise hierarchy. Initial focus has been on the top two levels: business planning and logistics functions layer (level 4), and the manufacturing operations and control layer (level 3).
The purpose is to provide consistent descriptive function terminology, information and operations models, as well as information data exchange. It does not currently address specific protocols or implementation.
It is comprised of multiple parts:
Part 1 describes models for integration and standard terms, definitions, functions, including product definition, production capability, production schedule, and production performance;
Part 2 describes object attributes that support elements in Part 1;
Part 3 describes manufacturing operational activity model and
Part 4 provides further details on manufacturing operations; and
Part 5 defines messaging requirements.
Parts 1 and 2 have been in the public domain for some time; Part 3, recently published, has significant importance to defining critical MES layer operations, such as tracking and tracing.
WBF: The Forum for Automation and Manufacturing Professionals (formerly called World Batch Forum) has built an implementation of the communication specification between Levels 3 and 4 based on XML, known as the business-to-manufacturing markup language (B2MML) for textual information exchange. Together, ISA 95 and B2MML intend to seamlessly coordinate business and production processes, ensuring least-cost interoperability.
Integration standards update
At the most recent ARC conference in Boston, initial meetings were held between OPC, OAG (Open Applications Group), ISA 95, B2MML, and MIMOSA (Machinery Information Management Open Systems Alliance). The aim of these meetings are to develop a framework so users can better understand each standard group, related standards, and when use of which standard is most appropriate, says David Connelly, president and CEO of OAG.
Connelly says he advocates a framework with TCP/IP as the base layer, OPC on the next layer, leading to a top layer consisting of options among the standards groups to be implemented by the user, based on application or industry relevance.
OAG is a relative newcomer to the world of automation standards, but it has a long history in ERP, as several ERP companies, such as Oracle, Marcam, IBM, and SAP, helped found it in 1994. This history makes OAG a potentially important player in these burgeoning enterprise integration discussions.
Connelly notes that OAG is a formal standards group, not a volunteer organization. Since OAG's inception, it has developed 434 schemas dealing with ERP integration standards, e-commerce, and specific manufacturing- related processes such as production, work orders, engineer-to-order, MES, and work in process.
David Greenfield, Control Engineering