Schaffhausen 2000 - Accuracy of Artificial Athlete - Rules of Standardization - Lab Accreditation

S01 ASTM Standardization System

The ASTM System for Consensus Standards Development

Ed M. Milner, Ed Milner Consulting Services, USA

Introduction

I have been personally involved as a member of The American Society for Testing and Materials ("ASTM") for more than 40 years, and find its methods of operation quite comfortable, although sometimes slow. I have depended on ASTM Methods for more than 50 years, in the synthetic resins, chemical solvents, synthetic fibers, and sports facilities and equipment areas.

It is not my objective here to debate the benefits of one method of standards setting against another, or to present a legalistic review of the process that is commonly used in the USA. Rather, what I hope to do is describe the system we have found useful under the presuppositions and assumptions that are basic in our culture. Presuppositions and assumptions are the meat of cultures, and I certainly realize that they differ from one to another.

Following the basic presentation I will be glad to at least attempt to answer your questions or entertain your opinions.

1) Why Set "Standards"?

1.1 Need for a common language for description of practices, terminology, and classification of materials and systems.

The advent of mass production methods, some 200 years ago, was made possible by the availability of interchangeable parts, having uniform dimensions and properties. The resulting economies of production efficiency revolutionized society. In the current age, the growth of international and multinational commerce makes such interchangeability a sine qua non. Unless we define and accurately describe product properties and dimensions today's international transfer of goods and services would grind to a halt.

1.2 Need for precision in description of materials and systems.

There was a time when measurements and tolerances were far less critical than they are today. I must admit that the USA's continuing dependence on English rather than metric units is a bit of a scandal. We are working on its resolution, but a massive system has a lot of inertia, and change is slow in coming. The "two unit" systems cost a lot of money to maintain, and lead to confusion. At least, we can agree to recognize the problems involved and actively work towards their solution.

1.3 Importance of mutual understanding between buyer, seller, and authorities.

A fundamental element of any valid contract is agreement between buyer and seller, the "meeting of the minds". Unless there are agreed upon ways to describe the items or actions in the contract such agreement on other matters is impossible.

2) Types of Standards

2.1 Company Standards

There is nothing inherently wrong with a given company or organization developing its own definitions, measurement processes, and units of measurement. It is limiting, however, if the group involved wishes to communicate with other organizations or businesses.

There is a valid concern that unilaterally developed standards include only the experience and insights of the originating organization. More broadly based standards and methods often give access to expanded knowledge of industry practices and the state of the scientific art.

2.2 Industry Standards

A given industry (and the beverage alcohol industry provides good examples) may develop perfectly usable standards in its own language. We think nothing of drinking "90 proof" whiskey, a descriptive term much less logical than "45% ethyl alcohol by volume". "90 proof" sounds like additional value when compared to "45 per cent" and percent by weight would require more alcohol than percent by volume. Even so, the public understands and accepts the industry standard description.

Industry-wide standards are a considerable step forward in commercial utility as compared to individual company standards. Their advantage, as well as their disadvantage, is that the development of industry standards requires competitors to work together - often, not an easy or simple task.

2.3 Professional Standards

There are well-established standards of ethical conduct among members of the learned professions. Systems for their promulgation and enforcement are well understood and observed.

In the US, such organizations as the American Society of Mechanical Engineers have done an excellent job of defining standards for such efforts as the design of pressure vessels used in the chemical process industry.

2.4 Government Standards

"Just let the government pass a law" is easy to say, but not necessarily the best, most scientifically valid way to set standards. Most countries have governmental administrative bodies that have the power to set rulings with the force of law. Such rulings often have only poor and inefficient appeals mechanisms. There is no tyranny like that of a minor bureaucrat defending his area of authority (and notice that I did not say "expertise")

Standards developed by legislative bodies are even more subject to political rather than scientific influence. Appeals to the body politic often reflect more emotion than cerebration.

2.5 Game Rules

The game rules promulgated by national and international sports governing bodies are in many ways similar to the company and possibly industry standards mentioned above. Every sport has its own dimensions for its field of play, and in some cases the properties of the surface used. Sports equipment rules cover dimensions, materials, shapes, and in some cases, performance.

The current work in baseball to define the performance of aluminum bats comes to mind. Researchers have been hired by the National Collegiate Athletic Association's Baseball Rules Committee to measure the performance of wooden bats and translate their findings into specifications for bats made of other materials. Olympic baseball is still played with wooden bats.

There has been a tremendous amount of work by fine researchers in the performance requirements for American football helmets. Unfortunately, their research findings led to a humanoid headform that, while it was a good analog for the human head, was not suitable for the routine testing required by helmet manufacturers. Manufacturers need be sure that helmets from a production line are both safe to use and comfortable to wear.

3) Some Processes for Standards Setting.

3.1 Unilateral

3.2 Governmental

3.3 Governing Body

3.4 Concensus

Each of the listed processes has its advantages and disadvantages.

Those set unilaterally by a company or governing body may be drafted quickly, but are limited to the input of the developer
Those set by government are not subject to argument or discussion, but must be obeyed.
The governing body of a sport may have the very best of intentions, but may not be aware of the limitations applying to those who attempt to meet its desires.

The aim of "Consensus Standards" is to include the inputs from all who have valid contributions to make. Consensus Standards are not easy to develop. They take time, discussion, participation, and, sometimes even rather heated debate. The end result, however, offers the best opportunity that I know of to include all the valid and useful information that applies. When the consensus process is properly applied most of the standards-setting pitfalls can be identified early in the process. The result is more efficient and precise standards for the use of all parties involved.

4) The Key Concepts for "Consensus Standards".

4.1 There tend to be many different "interests" involved in the development and use of a given standard.

Consumers, regulators, competing manufacturers, researchers, scholars and the general public may all have specific and differing interests in a given issue. The only honest way to deal with these differences is to recognize them and let them compete in the world of ideas. Open recognition of "interests" or "positions" protects both the participants and the public from unfairly drawn standards.

4.2 Each "interest" deserves to be heard.

4.3 No single interest should dominate the process.

No single organization has more than one vote except in the task group development phase of the work.

4.4 Each participating interest should be openly identified.

There are no anonymous participants

4.5 Standards should be written in such a way that their meaning and application are clear to all involved.

• Standard methods of testing should give reproducible results from lab to lab.
• Standard methods of testing should have well defined statements regarding precision.
• Standard methods of testing should not unfairly favor one material or system.
• Standard methods of testing should clearly state their area of applicability and the limits thereof.

5) The ASTM Process

5.1 Who/What is The American Society for Testing and Materials?

5.1.1 Independent Organization

As of the end of 1999, ASTM had 30,932 members. There were 20,391 Main Committee members, on 129 Technical Committees. There were 1,992 Subcommittees, the basic technical working groups of the Society. The society was organized in 1898.

5.1.2 Private Sector

ASTM is principally funded by the sales of its 73 volume "Annual Book of Standards", along with sales of copies of individual standards, special technical publications, symposia proceedings and other documents. ASTM does not maintain or operate any laboratories as such. All test development work is done in member laboratories or those contracted with by individual committees.

5.1.3 US Government Recognition/Support

5.1.3.1 US Government employees may participate as appropriate.

5.1.3.2 US Government employees hold no "special" membership or leadership status.

5.1.3.3 The US Government, by law, recommends use of private sector, consensus standards wherever possible.

Public Law104 -113 "National Technology Transfer and Advancement Act of 1995" sets forth the principles for participation in Voluntary Consensus Standards organizations by federal government employees. "The act is intended to reduce to a minimum the reliance on government unique standards, with the following goals:

• Eliminate the cost to government of developing its own standards.
• Provide incentives and opportunities to establish standards that serve national needs.
• Encourage long-term growth for U.S. enterprises and promote efficiency and economic competition through harmonization of standards.
• Further the policy of reliance upon the private sector to supply government needs for goods and services."

5.1.4 Legal Status

5.1.4.1 Antitrust Issues

Where competing suppliers and consumers meet and work together, care must be taken to be sure that antitrust laws and rulings are observed. ASTM Staff are trained to be highly vigilant in this area.

5.1.4.2 Standards hold "quasi-legal" authority.

A great many ASTM Standards are incorporated in governmental regulations at all levels, as, as such, carry legal implications.

5.1.4.3 Role of ASTM Staff

ASTM staff personnel assigned to Technical Committee meetings serve as advisers and consultants regarding standards-setting procedures. They carry no responsibility in the committees' areas of technical responsibility.

5.2 ASTM - ANSI - ISO Relationships

• ASTM is a "standards development agency".
• ANSI, the American National Standards Institute, accredits "standards setting agencies". ANSI is the American representative to the ISO.
• ISO, the International Standards Organization, is a private, worldwide, voluntary standards agency, Its membership in composed of the 133 national standards agencies such as ANSI. The ISO has over 180 technical committees. Most ASTM Standards carry an ANSI number as well.

5.3 The ASTM Committee structure

5.3.1 ASTM Membership

5.3.1.1 Open to all - Voluntary

5.3.1.2 Only one vote per "interest", but a given organization may have several members, either on the same or different Technical Committees.

5.3.1.3 Member Classification

Producer
Consumer/User
Independent Expert
General Public

Classification may change depending on the topic or product/system under study. For example, the person representing the manufacturer of a sports shoe might serve as an independent expert on the cushioning properties of a surface, provided that the surface is manufactured by a company other than his own and he has technical expertise in the area involved.

5.3.2 Committee and subcommittee membership is by request of the individual.

5.3.3 Task Group membership is by assignment of the Subcommittee Chair.

Task Group Members are assigned to task groups on the basis of their technical competency, availability of suitable laboratories, and interest in doing the work. Subcommittee and task group working papers are confidential, for internal use only.

6) How development of a Standard Method of Testing normally flows:.....(the "adjectives")

6.1 Need recognized, assigned to Main Committee.

New projects often result from letters directly to ASTM headquarters, which are then assigned to cognizant Technical Committees ("main committees").

6.2 Main Committee assigns to appropriate subcommittee.

Subcommittees are the working groups for given areas of technology or product application.

6.3 Subcommittee organizes "Task Group" and assigns project.

Subcommittee projects are assigned to ad hoc "task groups" competent to do the appropriate methods development or modification work.

6.4 Task Groups develop their plan of work and begin to develop or choose test procedures.

Task group members tend to do most of their test development work in their own laboratories They use available technology where possible, but may develop specific test procedures where necessary or advantageous. Task groups meet periodically to review their results and plan for next steps. When a method takes shape, they draft a proposed "test procedure".

With approval of the Subcommittee, the Task Group conducts interlaboratory tests of the new procedure. Following preliminary interlab comparisons, they will adjust the procedure or its written description as needed. Their objective is to produce a written procedure in such form that any competent laboratory technician can read the method, run the procedure, and obtain results comparable to those obtained by any other competent lab.

Once the procedure reaches this stage, the Task Group will organize and conduct formal "round robin" tests on matched samples to determine lab to lab agreement, accuracy, precision, and reproducibility data.

Following successful "round-robins, the Task group will submit the resulting method to Subcommittee for its approval. At this point the Subcommittee will take a "Classified" vote regarding approval of the method.

Following Subcommittee approval, a new method is submitted for "Main Committee" vote. Main Committee and "Society" votes may be taken concurrently. The "Society Approval" vote involves listing in the journal "Standardization News", so that interested parties not on the committee have an opportunity to call for the method, review it, and vote if they deem it important to do so.

7) The Voting Rules

7.1 "Classified Votes"

• All "interests" must be identified.
• Each "interest" can submit only one vote from its designated voting member.
• 60% of ballots must be returned.
• A member may vote "abstain", or "abstain with comments" if he deems it appropriate and still count in the 60%.
• Among the interests voting, producers must be less than a majority of the group.
• Users, independent experts, and the general public voters must constitute a majority of those returning ballots.
• A minimum of 90 per cent of the votes returned must be affirmative,
• All negative votes must include a written explanation of the voter's objections.
• All negative votes must be resolved.
• Each negative vote will be reviewed at the level (Subcommittee, Main Committee or Society) at which it occurs.
• The ASTM Committee on Standards is the final authority as to voting procedure.

A negative vote may be found "non-persuasive" at any level, but will be reconsidered if resubmitted at a higher level. Often negative votes are resolved with relatively simple editorial changes, but a persisting negative position may result in the method being run back through the development process.

7.2 "Non-Classified" Votes

"Non Classified" votes occur at the Task Group level only. Here participants are those individuals actively engaged in the development process itself.

8) How a typical "Standard Specification" might be developed. (the "adverbs")

8.1 Basic flow same as "Standard Method of Testing".

8.2 Includes selection of appropriate test methods.

8.3 Includes specification limit values.

8.4 All subcommittee and main committee votes are classified.

9) Does the Process Work?

In a word, yes. Is it slow and sometimes frustrating? Yes! A major advantage of the consensus standards development process, however, is that, in the development effort, consensus does evolve.

Those intimately involved in the effort to develop a standard gain a full understanding of its nature. They are then in an excellent position to explain both the nature and benefits of the resulting method to both Subcommittee and Main Committee members.