R1, later renamed XCON (for eXpert CONfigurer), was a rule-based expert system written by John McDermott of Carnegie-Mellon University to configure orders for Digital Equipment Corporation’s VAX computer systems. McDermott described the system in his 1980 paper “R1: An Expert in the Computer Systems Domain,” presented at the first AAAI conference; this entry uses the August 1980 conference date. The paper is in the official AAAI archive.
Configuring a minicomputer order was a fiddly task: a customer’s chosen components had to be checked for compatibility, and the right cables, cabinets, and supporting parts had to be added so the machine would actually work when assembled. R1 captured this expertise as production rules, hundreds of them at first and thousands later, and applied them automatically to each order. Built in the OPS5 rule language, it went into production use at DEC around 1980.
XCON mattered because it was one of the first expert systems that clearly earned its keep in a real business. It reduced configuration errors, sped up order processing, and was widely reported to save DEC millions of dollars a year. That visible payoff helped ignite the commercial expert-systems boom of the early-to-mid 1980s, when many companies rushed to build their own rule-based systems.
Why business readers should care: XCON is the proof point that expert systems could deliver real return on investment, not just laboratory demonstrations. It also previews the limits that later bit hard: the rule base grew large and expensive to maintain, and the enthusiasm it helped spark eventually outran results, contributing to the late-1980s AI bust.