May 6 2004
Professor Richard N. Langlois
Competition through institutional form: the Case of the Cluster Tools Standards.
Another excellent paper from Professor Langlois. This paper discusses the history of the Semiconductor manufacturing equipment and how it has evolved, particularly around the intense competitive pressures from Asia. Comparing a high tech industry to a commodity business may bring criticism. I would re-assert what
Matthew R. Simmons states that the energy industry is second only to the space industry in terms of the application of science. This paper's analysis focuses on the appropriate split between the firm and market, a similar discussion to what I am carrying out here for oil and gas. A discussion that focuses on the Joint Operating Committee (JOC). Langlois takes this discussion to its ultimate objective, that of course being the organizational structure as a competitive weapon.
"Industrial economists tend to think of competition as occurring between atomic units called "firms." Theorists of organization tend to think about the choice among various kinds of organization structures - what Langlois and Robertson (1995) call "business institutions. But few have thought about the choice of business institution as a competitive weapon." pp. 1
Key to this discussion is the standards used within an industry. Within the energy industry their has been a move to standardize the data elements between producers and suppliers. This has primarily been done through the efforts of the
Public Petroleum Data Model (PPDM) I have discussed here before. As a result standards in the energy industry have been established and are used by many of the engineering and geo-technical applications in this industry. This project is committed to the PPDM model as it is both international and most of the remaining work to be done is to resolve conflicts within the industry data elements. In Semiconductor's standardization was available but not used by everyone. Many of the larger "fab" manufacturers were able to establish and use their own methods of production. Through the evolution of the industry this made for some interesting positions being taken within the competitive marketplace. The semiconductor's dynamics of this structure and the changes to the marketplace as a result are interesting from the point of view of the discussion here on energy.
"Rather than a battle of the standards, the current situation might best be thought of as a battle of alternative development paths" the closed system of Applied Materials, with its significant internal economies of scale and scope, and the open modular system of the competitive fringe, driven by external economies of standardization. At this point, the forces favoring the integrated development path are more-or-less evenly balanced against the forces favoring the path of technical standardization. I analyze these forces in terms of the trade-off between the benefits of systemic innovation and systemic coordination on the one hand and the benefits of external economies of scope and modular innovation on the other Although standards have so far kept the competitive fringe in the ballgame, modularity in the industry may ultimately take a different, and somewhat more familiar, form, as some of the larger firms adhering to the standards become broadly capable systems integrators who outsource manufacturing to specialized suppliers of subsystems." pp. 1
As most people are aware the development of the semiconductor industry has been seen as a critical point in the development of nations. Here Langlois brings in a study from Berkeley to lay the groundwork of this discussion.
"In one of the few contemporary academic examinations of this industry, a study by the Berkeley Roundtable on the International Economy concluded that;
"... with regard to both the generation of learning in production and the appropriation of economic returns from such learning, the U.S. semiconductor equipment and device industries are structurally disadvantaged relative to the Japanese. The Japanese have evolved an industrial model that combines higher levels of concentration of both chip and equipment suppliers with quasi-integration between them. whereas the American industry is characterized by levels of concentration that, by comparison, are too low and [by] excessive vertical disintegration (that is, an absence of mechanisms to coordinate their learning and investment processes) (Stowsky, 1989)" pp. 3
Langlois then continues on discussing how the products are made in the semiconductor industry. This discussion is critical as background for the discussion here in the energy industry. I will pick out some of the stronger parts and make application to the energy industry later in this entry. Through analogy Langlois draws a litre of milk to the methods of keeping the semiconductor clean room. Instead of having everything "clean" for processing, and just as you would not refrigerate the entire house to keep the milk cold, the clean rooms were adjusted in size leading to new economies of scale.
"Indeed, there has arisen something of an international division of labor in the industry, partly by default. We can think of the more that 500 process steps in semiconductor fabrication as grouped into three phases akin to the steps in photo developing." pp. 5
"Instead of thinking about refrigerators, think now about dishwashers, and consider the problem of washing a kitchen full of dirty dishes. Using a dishwasher is a batch process; washing by hand is a continuous process. Loading the dishwasher may ultimately have a larger "throughput", but the first clean plate reaches the cupboard more quickly with hand washing. Batch semiconductor processing is like running dishes sequentially though many different dishwashers with many different capabilities. This creates a queuing problem, and the wafers must often sit around in WIP inventories while waiting to form a batch of the appropriate size for the next process step. By contrast, single wafer systems push only a single wafer through at a time (putting aside parallel processing steps), but the progress of that single wafer is not slowed as much waiting for other wafers to be ready." pp. 9
"Introducing a single wafer step into a batch fab instantly creates a bottleneck, of course, since throughput of the fab is limited to the throughput of the single wafer step. The obvious answer is to replicate the bottleneck stage in a parallel processing configuration. The need for parallel processing was the original motivation for common platform cluster tools." pp. 11
"Instead of running the same process in all four chambers, one could instead run different processes, using the wafer handler to moved the wafers from one to the other within a controlled atmosphere. This was the genesis of the integrated cluster tool which represents a genuine move in the direction of single wafer processing. The parallel configuration offers the benefit of redundancy, and can generate higher throughput when downtime is an issue; but as tools become more reliable, the serial configuration - which boasts superior cycles times - gains the advantage. (Lopez and Wood 2003)." pp. 12
"One way to marshal the necessary capabilities is within the boundaries of a single firm large enough to possess and wield all, or at least most of, the competences necessary to produce a cluster tool. Another way is somehow to organize and integrate through contract the competences of a number of distinct firms. The American semiconductor equipment industry uses both of these approaches simultaneously." pp. 12
"Applied (Materials) has quite deliberately chosen the opposite strategy - to develop internally capabilities in all areas of semiconductor fabrication technology. Initially, Applied did contract with firms like Peak Systems for an RTP module and GaSonics for a photoresist stripping module. Both of these arrangement generated contractual problems and were abandoned." pp. 13
This last point, "contractual problems" were earlier discussed
here. This is therefore a major demarcation in this analysis of energy in comparison to semiconductor industry. The ability of the market to define the contractual basis within a market system, I have suggested exists in oil and gas. This is the cultural framework of the industry that drives the financial, legal and operational decision making frameworks in oil and gas, and the direct mechanism to operate in oil and gas is the JOC. This is also the point of conflict that I am experiencing with the CIO's of the CAPP CIO committee. The market is attempting to provide a solution, being this discussion, to solve the administrative issues of the industry. Whereas the CIO's focused on the firm are breaking down what they need to such a fine level, that $50,000 is a reasonable level of contractual requirement. Work within the industry needs to be done, and that is consistent with the understanding of Langlois, and the CIO's have failed to see this point. I would refer back to this
table to see the makeup of the industry should be from my point of view. With the PPDM standards in this industry, any attempt to build systems that do not recognize the JOC would be a continuation of the forces necessary to micro manage all aspects of the industry by the individual firm. This is the reason that I attribute the CIO's are failing, and will continue to fail until they address these points.
"The emergence of standards."
"The process by which standards emerged in the cluster tool industry is rather different from those of well documented cases like the QWERTY keyboard, the VHS videocassette recorder, the IBM compatible personal computer or the 33 rpm LP record. In all of those cases standards emerged through or competition or "battle of the standards" among alternatives originally offered as proprietary schemes. In cluster tools, however, a single standard emerged immediately out of collective action with a fragmented industry." pp. 14
This last point showing a consistency between the energy and semiconductor industries. For these purposes here, there is only one data standard, PPDM in oil and gas. The value of which is well articulated by Langlois in the following two quotations. I would also add to this discussion of data standards, the copyright that I hold in the development of this research. As I have asserted I will only license one firm to develop the software for these purposes in order to focus the energy industries efforts on the one right solution. Using a shot gun approach to system development will not work with such a high level of technology being employed.
"For the moment, then, both development paths seem to be surviving, and neither is obviously driving out the other. Why? Let us pause to think about the basic economics of closed proprietary systems versus open modular ones. The primary benefits of a closed system lie in the ease of systemic coordination and reorganization. When the nature of the connections among the elements in a system are changing or idiosyncratic to application, a unified organization can more cheaply coordinate and fine tune the connections. The value of such systemic coordination depends on both technological and demand factors. In some respects, and in some technologies , the value of idiosyncratic systemic coordination may be exogenous. In the automobile industry, for example, some degree of "integrality" may be inherent in the nature of the product (Helper and McDuffie 2002) Moreover as Christensen and his coauthors have argued, an integrated organization is better able to fine-tune product characteristics to achieve greater functionality in an environment in which users eagerly demand such functionality. (Christensen, Verlinden, and Westerman 20002)" pp. 18
These are the benefits that accrue to the energy industry as a result of one software solution being built here. The focus is not diluted through the needs of venture capitalists and competition among rivals. The focus is on providing the best software solution possible to the industry in the manner that the industry defines. Langlois then shows the benefits of an open modular system. This open modular system is what I am proposing for the "market" of the energy industry. An "open" market comprised of many Joint Operating Committees.
"On the other side of the ledger, an open modular system can more effectively direct capabilities toward improving the modules themselves (Langlois and Robertson 1992). Such a system harnesses the division of labor and the division of knowledge, allowing organizational units to focus narrowly and thus deeply; at the same time, it magnifies the number of potential module innovators, and thus can often take advantage of capabilities well beyond those even a large unitary organization could marshal. In this way, an open modular system "breaks the boundaries of the firm." There are both static and dynamic benefits. At any point in time, a user can "mix and match" components from a wider variety of sources to fine-tune the system to his or her taste, and thus reach a higher level of utility or tailored functionality than pre-packaged system could offer. In the semiconductor equipment industry, this is called "best of breed."" pp. 19
"More important perhaps are the dynamic benefits. Over time, an open modular system can lead to rapid trial and error learning and thus evolve faster that a closed system." pp. 19
This is how I foresee the energy industry developing greater focus and capabilities based on innovation. The organizational structure of the JOC provides the ability to deal with the market as it needs. The market evolving essentially in response to the producers needs. This needs to be defined and built within a software structure that is the purpose of this blog. When we add compliance to the mix, I fail to see how the energy industry could continue on without the re-organization happening first, and of course that won't happen until such time as this software is operational.
Technorati Tags: Genesys, Langlois, Capabilities, Firm, PPDM, Research, Organization
