Innovation in complex systems? Oh, FFS…. And CCS

By FFS I mean “Full-Flight Simulators”.  What am I on about? So, innovation in mass produced commodity products (aka “widgets”) is, cough, relatively straight-forward.  Lots of opportunities for iteration, incremental learning, process and product innovation, tacit knowledge creation/management.    Shakeouts after the establishment of a ‘dominant design’, followed by incremental shifts that squeeze a leetle more efficiency out … But what if your product is incredibly complicated, and essentially a “one” (or maybe up to five) “off”?  Say, for example, flight simulators.  That is, what if it is part of a “Complex System”  (CS).  You see, I’ve been reading, awestruck at its coolness, this –

Reading and loving – Miller, R. Hobday, M, Leoux-Demers, T. and Olleros, X. 1995. Innovation in Complex Systems Industries: The Case of Flight Simulation. Industrial and Corporate Change Volume 4, Issue 2 363-400

Miller et al. argue that

while the conventional model may apply to mass market commodity products it is highly unlikely to apply to another important group of products and industries, classified here as CSs.2 As Part I argues, CSs account for a significant proportion of industrial output. In contrast with commodity goods, complex product systems are large item, customized, engineering intensive goods which are seldom, if ever, mass produced.3 Examples include flight simulators, telecommunications exchanges, electrical power equipment, military systems, airplanes, helicopters, flexible manufacturing systems, chemical process plant, intelligent buildings and nuclear power equipment.

(Miller et al. 1995: 364)

CSes have three ‘watch out, these aren’t just widgets’ features . They are

“first, they made up of many interconnected, often customized, elements (including control units, sub-systems and components), usually organized in a hierarchical way; second, CSs exhibit non-linear and continuously emerging properties, whereby small changes in one part of the system can lead to large alterations in other parts of the system; and third, there is a high degree of user involvement in the innovation process, through which the needs of the economic environment feed directly into the innovation process (rather than through the market as in the standard model).”
(Miller et al. 1995: 368)

They did a bucketload (like 120!!) of interviews and really got “into” the development of the flight simulator industry, and tell the tale well –

FS was born when Ed Link patented a simple mechanical flight trainer in 1929- During World War II electronic analogue simulators were built to train pilots and reduce the number of accidents. Early commercialization began with Link, Miles and the Wright Brothers. In 1951 Redifon (now Rediffusjon) built a Stratocruiser simulator for BOAC. BOAC and Lufthansa placed initial orders with CAE of Canada in the early 1960s.

From the early 1950s to the mid-1960s (prior to digital computing) a long period of experimentation took place, but there was little in the way of landmark innovations. Analogue computers improved gradually, as did the hydraulics and visuals. During this period the industry began a slow take-off.

During the late-1960s digital mainframe computers took over from analogue ones, leading to a rapid improvement in the fidelity, speed and capacity of FSs. However, up until the late-1970s pilots were mostly trained in airplanes. Simulators were viewed as a complement to live training rather than a substitute for it. Some training credits were granted by the regulators, but the process of certification was ill-defined and informal. Simulator technology was perceived as inadequate for manoeuvres such as take-off, landing and missed approach. Increasingly, though, the needs of more powerful jet aircraft encouraged a focus on problems such as air turbulence, recovery manoeuvres and landing and take-off procedures so that costly and dangerous live training in aircraft could be minimized.
(Miller et al. 1995: 376)

It’s a complicated business of course –

Full-flight simulators (FFSs) are full-size replicas of specific aircraft cockpits. They combine mathematical models and original flight data to simulate the behaviour of the aircraft and record pilots’ responses to changing conditions. Given the cost of commercial flight time, pilot training and re-training is carried out in FFSs.

(Miller et al. 1995: 377)

and so

FS makers are required to master at least four technical fields: (i) the skills to integrate interdependent hardware and software components (motion, visual, computer and cockpit) into a coherent whole (the simulator); (ii) the know-how to use and develop the mathematical simulations which replicate the behaviour of the aircraft (as well as the actions of pilots and crew); (iii) the detailed knowledge of client requirements for training, checking and quality programmes which involves theoretical work as well as teaching methods; and (iv) a knowledge of rules and regulations (notably the acceptance test guides) which specify the requirements for simulator approval.
(Miller et al. 1995:381)

So, given that, it comes down to a very complicated and “uncertain” set of processes, that go far beyond “the invisible hand of the market” –

To sum up, the need to coordinate innovation in FS called forth a complex institutional superstructure. New technology proposals are channelled through professional bodies such as the Royal Aeronautical Society. Acceptance test guides are established by regulators who then specify approval requirements and validate tests during and after the development of an FS. After contracting, trust and reciprocity are necessary between buyers and sellers. Because many uncertainties have to be resolved during the process of innovation in FSs, they cannot be purchased as arm’s length market transactions as in the standard model. Instead, intense relational transactions develop, allowing for constant information exchange and regular interaction between industry participants. Continuity of relationships is valued and respected, and helps define the competence of partners. Innovation in FS unfolds within a set of governing institutions where, as discussed below, cooperation and competition co-exist.
(Miller et al. 1995: 384)

So, Schumpetarians, this is a bit more complicated than you’d like to believe-

As noted in Part I, in the conventional Schumpeterian model, radical technological discontinuity leads to creative industrial disruption. Subsequent process and product innovations shape observed patterns of exit and entry (Tushman and Anderson, 1986; Utterback and Suarez, 1993). These elements of the conventional model do not fit the FS industry, nor are they likely to apply to other CSs industries (Hobday, 1994).
(Miller et al. 1995: 386)

And, generally, we forget the past, (if we ever knew it), and fill it in with convenient just-so stories…  Research like this reminds us that

“the institutional structures and processes taken for granted in today’s industry did not simply occur or arise out of market transactions. On the contrary, they were initiated and crafted by a small number of key individuals widely recognized across the industry as entrepreneurial leaders, not only in the field of technical innovation but also in the areas of regulation, standards and consensus building. Each successive wave of technological change was associated with one or more industry champions, including Edward Booth (Federal Aviation Administration), Captain Ray Jones (Royal Aeronautical Society), Brian Hamson (CAE), Vince de Paulo (American Airlines), Hans Dieter Hass (Lufthansa) and M. Bess (Air France). Drawn from a variety of groups in the innovation structure, these individual were entrusted by their organizations to bring about progress in the national and international decision-making institutions, for the benefit of the entire FS industry.”
(Miller et al. 1995: 390)

Nothing, absolutely nothing, in this makes me think that CCS ever stood a chance, as a world-wide diffused technology dependent on not just human smarts (FFS) but also co-opeative geology.

We’re so toast. Carpe the goddam diems.

BTW – Here’s the abstract

The paper proposes that the notion of complex systems usefully describes a group of large scale, customized products and their associated supply industries. Examples include flight simulators (FSs), telecommunications exchanges, military systems, airplanes, chemical process plants and heavy electrical equipment. Complex systems, made up of many interconnected customized components, exhibit emerging properties through time as they respond to the evolving needs of large users. Taking the FS industry as a case history, the study identifies some of the basic rules governing innovation in this industry. These rules contrast sharply with those typically found in the ‘conventional’, market contest Schumpeterian model. Innovation in FS is coordinated by an institutional structure made up of suppliers, users, regulators, industry associations and professional bodies. In contrast with co.tventional market selection, new designs are negotiated prior to product development. Long-term stability among FS makers is observed, despite radical technological discontinuities, as industrial adjustment occurs via the exit and entry of specialist suppliers. There is no dominant design in the usual sense, nor do the conventional rules of volume competition and  process-intensive innovation apply in FS. Competitive strategies remain focused upon design, engineering and prototype development, rather than incremental process innovation. Collaboration occurs among the innovation actors within institutions created by them to harness innovation and to allow new product markets to develop. Recognizing the limits of a single case, the paper suggests that other complex systems might exhibit similar processes for governing innovation and reducing risk and uncertainty in the  absence of conventional Schumpeterian market mechanisms.