10 CHAPTER 2 A Short History of Configuration Technologies

Nearly 40-years of configuration technologies motivated us to give this brief overview of the main configuration technological streams. We outline the technological developments in the field starting from the first expert systems of the late 1970s down to today’s configuration solutions. The vastness and intricacies of the configuration field makes it impossible to cover concisely all of its relevant aspects. For the purpose of this overview, we decided to focus on the following four different yet overlapping technological developments: (1) rule-based configurators, (2) early model-based configurators, (3) mainstream configuration environments, and (4) mass customization toolkits. 2.1 Rule-based Configurators Weak Artificial Intelligence (Weak AI) (Searle, 1980) denotes an interpretation of AI in the narrow sense: it supports the accomplishment of specific problem-solving tasks (such as configuration), not encompassing the full range of human cognitive abilities (see Russel and Norvig, 2003). In the line of weak AI (the term was not known at that time), many expert systems were developed in the 1970s. Examples of successful configuration expert systems are (1) R1/XCON (a first version was developed in the late 1970s), a computer configurator (Bachant and McDermott, 1984; McDermott, 1982; Soloway et al., 1987) and (2) VT (vertical transportation), a system for handling the design of elevator systems at the Westinghouse Elevator Company (Marcus et al., 1987). One of the major reasons for applying expert system technologies was the failure of procedural (e.g., FORTRAN-based) approaches to solve the configuration problem. As an example, we focus on the R1/XCON configurator. R1/XCON was an expert system initially implemented by John McDermott on the basis of OPS5 (Forgy, 1981); that is production rules were used for expressing configuration knowledge (see, e.g., Figure 2.1). R1/XCON was developed to support order generation for VAX computer systems (initially for Digital Equipment VAX-11 systems, later for PDP, MICROVAX, and MICRO-PDP). Before R1/XCON was introduced, computers at Digital Equipment (DEC) were configured manually. As a consequence, sales personnel created erroneous, incomplete, and infeasible configurations, which caused enormous overheads and costs. The overall net return to DEC was estimated $40 million dollars per year as the result of the successful delivery of valid and completely configured systems.