by Charles Hoes

One of my main headaches during the past year or so has resulted from attempting to find a single way to meet worldwide product safety standards. One of the companies that I consult for has a product line of very large and expensive tools that are installed in a variety of countries around the world. For many years, the customers were located in three Asian countries: Taiwan, Korea and Japan. Because the tools are quite similar to semiconductor production tools (but much larger than the tools used for making computer components), we have been able to convince the customers and their regulators that complying with SEMI S2 (an industry safety standard developed for the semiconductor industry) was sufficient. However, that situation has changed because the new customers are located in a far broader spectrum of countries, including the United States, Europe and China. In addition, the tools have changed so that they are not quite the same as semiconductor tools; therefore, customers are not necessarily willing to accept SEMI S2 compliance as adequate.

Because of these changes, we now need to show compliance with European Union directives, various U.S. standards, federal and state OSHA regulations, and a host of other, often conflicting, standards and requirements. Some of the requirements allow (and, in fact, mandate) system safety analyses to be the basis for determining acceptability. These are almost all European Union requirements; the U.S. and other countries still have not embraced the system safety approach to product design and evaluation. Those requirements that do allow hazard analyses and risk assessments to be the basis for safety assessments are fairly easy to comply with globally because the hazards, risks and appropriate controls are related to the system, not to the country of use. The ones that are difficult to deal with are those (such as most U.S. regulations, like OSHA) that are prescriptive in nature. There are still many prescriptive standards demanding certain design details, rather than directly addressing the issue of risk. They make the assumption that if they specify how to build a product in enough detail, then the product will be safe.

In general, the detailed design requirements are intended to be adequate solutions to common safety problems. This is usually fairly easy to follow, as long as you can use and follow one set of codes or standards — or only sell products to one country. An example that comes to mind is in designing access to large tools using stairs, ladders, platforms and railings. There are many "local" requirements for these items, each of which is acceptable in some jurisdictions but often in direct conflict with detailed design requirements in others. Often, there isn't even a "worst case" or "most stringent" solution because there is no overlap in the design requirements.

Faced with this frustration, we finally decided to make a company-specific set of requirements that meets all of the known standards when possible, and selects something that we deem to be reasonable for those that conflict. This is a risky approach from the point of view of achieving full compliance because it is not possible for a single design to comply with all of the conflicting requirements. However, it might not be risky from the safety point of view because the chosen requirements were judged to be "safe enough" by whatever standards-setting body made that decision in the first place, and they were developed to be internally compatible.

I find this whole approach to be extremely frustrating and unnecessarily expensive. In our global economy, it is necessary to sell products to many different countries or regions. It is not acceptable to design special products for each region when each product is highly complex and expensive — indeed, a one-size-fits-all approach is needed. It is clear that various standards and codes achieve similar levels of safety with whatever approach they use. I personally dislike doing busy-work just because each organization (or country) has decided to take the "not invented here" approach of demanding its own solution.

I think the solution to this dilemma is to develop unified international standards that are based upon risk, rather than detailed design requirements. This provides maximum flexibility to achieve a safe design, without having to create a plethora of designs to meet the various local requirements around the world.

There is a growing effort to consolidate international standards to address this issue, and in many cases these new standards attempt to address the issues from a risk-based point of view. Unfortunately, many of the committees creating these new standards are composed of people from the "old way" of using prescriptive standards. They seem to like the idea of basing the standards on risk, but do not necessarily know enough about how to accomplish this to put it into the wording of a standard. There is a strong tendency to carry over the prescriptive approach, but using the terms of risk-based safety. This can (and often does) result in standards that are neither fish nor fowl. They are really prescriptive design standards, but are not system safety-based standards either. Unfortunately, the result of this is often a situation where the standards need to be interpreted and those interpretations are often provided by commercial entities who wish to make the acceptable solution the one that they happen to sell. This happens because you can't really follow the standards as written; there are often no provisions for determining what is "good enough."

I believe that we, as system safety practitioners, need to take a much more proactive role in ensuring that conflicts do not exist between standards, and that they are as compatible as possible given differences in such things as voltage and frequency of electrical power in various countries. We need to ensure that professionals who are knowledgeable in the field of system safety are included on the standards committees, rather than sit back and hope that the committees somehow make the transition from prescriptive design standards to risk-based standards on their own. There is a huge paradigm difference between the two approaches to safety, and those who are in the prescriptive mind set have no way of even knowing that their view is incomplete (which is the tricky thing about paradigms; they are so complete that they don't allow a person to understand how limiting they are). In my opinion, we continue to do our customers and our employers a disservice by not being more vocal in expressing the view that risk should be the focus of safety efforts, not compliance with design standards.