Interoperability Barrier No. 1: The CIO

June 12, 2008
According to AMR, 60% of Most IT Budgets Are Spent Attempting to Deliver to 2002 Expectations

Read Charlie Gifford's blog... Hitchhiking Through Manufacturing

By Charlie Gifford, Contributing Editor, 21st Century Manufacturing Solutions LLC

At the MESA 2007 North American Conference, former AMR Research senior analyst Colin Masson conducted the solution-provider CIO Panel with representatives from GE Fanuc Automation Americas, Rockwell Automation and Siemens Energy & Automation. Masson asked the panel members who among them knew of ISA-95 or B2MML and was planning to standardize for their B2M interoperability using one of them or another manufacturing operations management (MOM) standard. No one on the panel knew what what any of these standards were.

The panel’s consensus strategy for interoperability was to use their vendors for ERP or enterprise interface application (EIA) middleware to do services-oriented architectures (SOA) and B2B and B2M interfaces and transactions. Sounds sexy, but it’s a wrong-headed approach.

The entire room of end users, consultants and salespeople from the MES solutions space were simply floored. All three vendors on the panel claim some level of compliance to MOM standards, but none of their executives endorsed or had any concept of open-standards-based interoperability between their MOM systems, suppliers or customers. Unfortunately, this CIO panel’s mentality is typical of the executive and mid-level management in manufacturing companies and their vendors.

Why are these CIOs (and their counterparts) THE No.1 barrier to making adaptive manufacturing a reality in North America?

Besides the factors of age, budget and ego, four paradoxes play a role.

The CIO Paradox: Most CIOs have spent millions on the Y2K transition only to install monolithic, highly customized systems based on custom information models and point-to-point interfaces. The systems have a high cost of ownership, and the architectures do not scale globally to enable a pull network of suppliers and customers, so the cost of any change is high.

According to AMR, 60% of most IT budgets are spent attempting to deliver to 2002 expectations. The manufacturing markets of 2008 have made the Y2K architecture and investment obsolete.

The CIO’s budget has never been tighter, while his department is moving into doing multiplant manufacturing operations systems on a shoestring. The CIO is risk-adverse and does not want to support another IT standard at the very time when the evolution of open semantic information and business process transactions models will lower his cost and risk significantly for the first time.

The Vendor Paradox: The leadership at software companies does not want to drive customers towards open interoperability, application information models or standard business process transactions. Why? Because they want to control and protect their customers through their proprietary terminology, business processes and message structures. Truly, they are not concerned about lowering cost of ownership, change or interfaces, no matter what they say in their marketing brochures. They make money on the custom interfaces, analytics and reports, not configurable, open systems.

As a system integrator, I did $25 million in MOM projects of which $10 million to $12 million was doing customer backside applications to vendor “solutions.” The unspoken policy is “Keep the CIO and his staff in the dark about open MOM standards with vaporware discussions on Lean Six Sigma and SOA. Never explain that Lean IT is really applying the MOM open standards as “the semantic standard work for SOA.” IBM is the only large IT vendor who has open realized that its business process mapping tools, SOA and data warehousing solutions cannot succeed without standards-based semantic information models as their bases.

The Industry Analyst Paradox: The vendors are the analyst’s bread-and-butter clients, so industry analysts with few exceptions do not aggressively make or endorse the open MOM interoperability business case to the end-user executives so they can in turn force their vendors to adopt and support the efforts.

Analysts are not addressing “the end-user crisis” on open standards by strongly encouraging them to contribute to open MOM standards development efforts. They do not explain that the final results of these standards, once adopted at any level, will drive or limit the form of their business models, SOAs and business process adaptability. Except for one analyst at AMR Research, none participate in the development of any of the MOM standards; they do not even review and comment on the drafts. Except for AMR, none of them are doing annual briefings with the MOM standards groups; but all are providing their uninformed assessments on the maturity or applicability of the standards to the CIOs and their staff. In fairness, ARC Advisory Group did address briefly interoperability a few years ago.

The Professional Society Paradox: Why are there only 150 people, not 500 people, at the North America conferences for WBF and MESA, but analysts’ meetings get 500 people? Because CIOs would rather have vendors and analysts tell them what to do than recognize their own responsibility to build internal competency by seeking out the experts in adaptive manufacturing systems. Where is the leadership? It is in Europe and China.

The professional societies that develop and support automation and MOM standards also lack the executive contacts and relationships to make the business case for their interoperability products. Society leaders at ISA, MIMOSA, OMAC, OPC, WBF and others do not understand how make a business case for their own products to their customer executives. OAGi and AIAG are notable exceptions.

I am part of the effort to form the Industrial Interoperability Compliance Institute (IIC) within the Automation Federation. I will not go into detail about the IICI here, except to speak to my direct experience with the lack of recognition by corporate IT departments for the solutions required for adaptive manufacturing.

The Open O&M standards set is the adaptive manufacturing solution and application framework for providing the information models for interface and application data models and transaction models. All of the Open O&M standards (OPC-UA, ISA-95, ISA-88, OMAC, MIMOSA and OAGIS) have individually proven use cases for their part of the adaptive manufacturing application framework, but their successes have largely been due to a small group of innovative people inside of a small group of innovative manufacturers, consultants and vendors.

I get disagreement about this because each standard has widespread traction at the plant level in one or two industries, but not at the executive levels. This is proven by the fact that all of the standards committees have very little end-user participation now due to downsizing of engineering groups (except for the occasional visionary). Corporate IT simply does not realize the need to participate actively or to acquire production process competency at all.

CIOs simply are looking to their ERP, automation and supply chain vendors to tell them how to run their companies’ plants and integrate their vertical business models. This is utter stupidity that is increasing the “sophistication gap” between North America, Asia and Europe.

A new information technology has developed that actually provides CIOs with one of the primary tools for adaptive manufacturing. North American CIOs still look to their vendor for the semantic and transaction models for SOA; the software vendor provides one, but the information model is not openly interoperable with any other vendors’ models.

Custom SOA is actually very expensive to own due to the high cost of change and the barrier to adapting business processes to new and changing global markets. There is no open interoperability between applications.

Now, as these CIOs push for “Global Lean Supply Chains,” the problem magnifies by 1,000 times, since all of their suppliers and customers now have custom B2B information models and transactions, even if they standardized. Some CIOs were forced to see this light when WalMart and Department of Defense required EPC RFID compliance of their suppliers. But the CIOs did not bring this interoperability lesson back to their companies.

What is really sad is that these same IT departments from manufacturing companies are not even participating in developing the ISA100 industrial wireless standard and are letting the vendors write the standard for them. So while the IT departments and their leaders have been drinking their vendors’ KoolAid, the group of dedicated visionaries of end users, consultants and open-system vendors has been building the interoperability parts for the adaptive manufacturing framework for applying an SOA globally.

There is not a better solution available for maintaining competitive advantage in the 21st century for a manufacturer than the combination of applying OPEN O&M interoperability standards within an enterprise SOA. They bring common definition to data, interfaces, workflow definitions, resource specification, KPI metrics and financial metrics. Adaptive manufacturing is enforcing a common information model across all enterprise and plant systems, business process mapping tools, data warehousing and B2M transactions with suppliers and customers, providing clear communication to empower change events.

We need the industry analysts and end users finally to get involved in open interoperability to make the strong business case to CIO to support these revolutionary organizations and not their vendors’ legacy systems and custom SOA approaches.

Otherwise, learn to speak Chinese.

Read Charlie Gifford's blog... Hitchhiking Through Manufacturing