The dichotomy of open standards

Dichotomy of open standards

By : Jim Pinto,
San Diego, CA.
USA

The definition of a "standard" is simple: operating characteristics that everybody follows. Therein lies the rub. Someone has to be the leader, to develop the standard that others follow.

Conflicting objectives continue to cause endless debate. To help clear the confusion, we must understand that technology developers need to recoup their investment through one or more of the rules for open standards.

This article was published by:

Automation.com, March 2002

A version of this article was also published in December 2001 by

In the industrial automation business, everyone agrees that standards would make a life a lot easier - everything would work together. End-users continue to ask for interoperability as a means to achieve vendor independence. But, this the exact opposite of what all the primary suppliers want. Standards turn proprietary products into commodities, with lower profit margins.

The search for a fieldbus standard is a typical example of the futility of trying to set standards. The dream of a unified fieldbus never emerged simply because the vendors could never agree - each group of vendors pushed their own favorite. The international committees dragged the debate on and on, until several different industrial networks were approved as “standards”.

Definition of "standard"

The definition of a "standard" is simple - operating characteristics that everybody follows. Therein lies the rub. Someone has to be the leader, to develop the standard that others follow.

The idea that committees composed of both users and vendors can set standards is futile. Users may perhaps be able to collaborate; but few can invest the time or expense necessary to drive a standard. And vendors can seldom agree - they are, after all, in direct competition with each other, each seeking to develop technology and market advantages over the other.

Who benefits from standards?

Standards provide openness and interoperability between products from different vendors, reducing products to the level of commodities. Hence, end-users are the primary beneficiaries. However, no single user is large enough, or strong enough, to demand and set horizontal standards.

GM tried to force a standard with MAP and failed. And GM continues to write "white-papers" on 1131 and similar proposed standards, but that's all they remain - white papers.

The intrinsic problem is this: Vendors want proprietary differentiation, which generates higher margins. Technology development is expensive and everyone understands the financial rules for expense amortization or write-off. Once the status of a standard is achieved (through technology and/or marketing leadership, or timing) other vendors will simply be required to conform - giving the clear proprietary (and financial) advantage to the owner, or consortium, that controls the standard.

Rules for Open Standards

The conflicting objectives continue to cause endless debate. To help clear the confusion, we must understand that technology developers need to recoup their investment through one of the following rules:

Rule 1
Licensing the technology. This may be through up-front fees for technology transfer, or per-copy sales of ASIC chips, hardware, software or firmware.
Rule 2
Making everything open and free, to expand involvement. The developer is far ahead on the learning curve and followers contribute to the leader’s leadership.
Rule 3
Introducing "free" open technology to combat the entrenched position of a dominant market leader.

Famous "open" battles

Consider some of the famous "open" battles. DOS became a standard (and made Microsoft rich) through the myopic munificence of IBM. IBM thought they'd sell PC hardware to catch up with Apple, but the hardware soon became a commodity. The only non-commodity was DOS - you had to pay a license to use that. IBM then thought they'd win with OS/2, but their previous largesse had already made Microsoft too powerful to catch and they got upstaged with Windows, which has now become almost a monopoly. It was a combination of timing (for Microsoft) and marketing mistakes (on the part of IBM).

There are several similar examples in the vast growth of the PC business - Unix and Windows; PC/AT bus and IBM PS/2; Novell and NT networks; EtherNet and ARCnet.

Industrial Automation examples

Some vendors "pretend" that they are simply giving away their technology freely for everyone to utilize when in reality they are encouraging others to use it as a standard. Example, Profibus: Siemens is the original developer of Profibus and continues to gain advantage through its proliferation (Rule 2). And, Siemens continues to sell Profibus chips (Rule 1). Meanwhile, the followers believe that what they support is indeed "free and open".

The Rockwell strategy in making DeviceNet and ControlNet "open" follows a combination of all three of the rules. They allowed other vendors to purchase or license their hardware, software and firmware (rule 1); they expanded involvement through gaining a bunch of followers, all eager to find new ways of connecting to the vast Rockwell (A-B) installed base of PLCs (rule 2); and, Rockwell was combating the encroachment of Siemens-dominated Profibus and Phoenix Contact sponsored Interbus into their US backyard (rule 3).

Exceptions?

There is one exception to Rule 2: The US Government (ARPA) really gave away the Internet - and caused a new revolution. Perhaps there is a lesson to learn there.

Return to Index of all JimPinto Writings
Return to JimPinto.com HomePage