For a topical opinion piece, I think the article holds up surprisingly well. The SEMI Sensor / Actuator standards are a reality. Many new 300mm semiconductor production tools have been designed around their capabilities. And yes, the market did pick the dominant solution. It was the same solution, by the way, that the task force attempted to pick in 1995. It certainly didn't have to be that way, but we had our finger on the key issue that probably decided the contest.

SEMI standards committees are busier now, with more new documents in process than ever before. The SEMI process is still not appropriate for making controversial choices. Efforts to standardize component interfaces for modular, surface mount gas systems, for example, have come to an impasse after several years of struggle.

The article is too pessimistic in its view of the limitations of the standards process, however. I now see how we might have worked our way out of the mess and made an informed technical choice. My experience was limited to large, consensus-based standards organizations like SEMI. As I have discovered, its not the only way.

Over the past ten years, hundreds of small, independent standards consortia have formed. Their strength lies in the ability to move very rapidly and to decide issues by majority vote. Any standards body is most effective when it includes all parties that have a stake in the issues. Developing consensus on major items is still key in the independent consortium. Large, disaffected players can still doom the resulting standard in the marketplace. But the group can have more independence from small, vocal minorities. Consider the issues, look out for the common good, take a vote, get on with business.

As with any historical "what if" exercise, its hard to say whether the sensor bus standards would have been born sooner and with less pain and suffering under a different organizational structure. There were times when we had an overwhelming majority of the votes, however, but we couldn't move. MFC companies played a major role in the initial development of the sensor bus. Tiny as we were, I don't think we could have financed an independent standards operation. We could have kept our heads down and finished a low level standard by 1993, however, as I suggested in the original article. Things would have turned out differently, I'm sure. The companies that had a technology edge in 1993 did not have the equivalent edge in 1998.

If you are frustrated by the limitations operating within of large, consensus-based standards organizations realize that there are alternatives. If a timely standard is good for an industry, manufacturers and customers alike, then that standard should exist. Bringing that standard into being proactively and pre-competitively is so much the better. Waiting around for the dust to settle does not always produce an optimal result.

As it turned out, market conditions delayed the commercialization of 300mm process tools long enough that the sensor bus standard was ready when needed. Its generally poor execution, however, making the market window because it stayed open longer than you thought.

David Sheriff

August, 2000