The ILS project dealing with Total Quality Management (TQM) was called QED. This was an early example of what later became known as a "performance support" system. It was intended to assist "quality teams" as they identified and solved problems using a structured problem-solving framework. The framework identified the steps in problem solving, and our system was intended to provide tools and support for each step, as well as context-sensitive help and advice, and a case library of problem solving expertise that would assist users in working through their problem solving projects.
The QED project was done at the behest of Ameritech, one of the ILS corporate sponsors. Unforturnately, the project had a short run and was cancelled when Ameritech decided to forsake TQM in favor of "Breakthrough Leadership".
We heard many comments from the workforce, during the course of the QED project, about how TQM would be another "flavor of the month" management initiative. We were concerned about it, and this assessment was presented to Ameritech management -- and downplayed by them. Yet, it turned out to be prophetic. We await the fate of Breakthrough Leadership.
The QED team had a wrapup pizza celebration in the apartment of Debra Jenkins (an Ameritech intern at ILS), and talk turned to the lessons we had learned in trying to implement a system like QED, which was intended to both teach problem solving and assist with problem solving within a particular framework: a tricky combination. We concentrated on the idea of teaching problem solving, and struck on the idea of an educational game and exploratory environment that would support "detective stories" (and thereby, logical deduction in a "real world" context).
We imagined implementing a space, and then populating that space with characters: characters at the crime scene to be interviewed, and other software agents who would represent the coroner, the evidence technicians, and so forth. The environment was to be relatively "low-tech", with a few scientific instruments and other "manipulables" but mostly implemented in terms of simple software agents who had personal knowledge of various events, and who would discuss things if the right questions were asked. The technology was to be little more than "agents as Ask Networks" with a few additional wrinkles (mostly conversational dependencies) to ensure that students followed a path of discovery on their way to a solution.
We spent a fair amount of time, that day and later, working out the details of a problem solving situation involving an apparent food poisoning in a restaurant -- Colette Marine suggested the original conception and devoted considerable time to this, and did a stellar job. Our conclusions were encapsulated into a short project proposal that was circulated for comment in the spring of 1992.
Debra Jenkins finished her program of study at ILS, graduated with a Masters degree, and went back to Ameritech in the New Products group. Soon she was orchestrating the details of a new Ameritech/ILS project to coincide with Ameritech's desire to provide all manner of networked services and "video dial tone into the home".
A range of proposals was generated by interested parties at ILS, inluding one that Ray Bareiss and I wrote. It was the same as the idea circulated for comment in the spring of 1992, but Ray wanted to make it an "environmental" detective, so we put it in that way. A meeting was held, the idea of educational games was endorsed, and the first game to be implemented was ..... a game to teach retailing, later to become a GAMES game known as Sell.
Only one GAMES game was implemented at ILS:
In the summer of 1995 an Andersen intern informed me that I was being assigned to the Covey/ProListen/Pinnacle project. The idea here was to create an environment where AA&Co. consultants could practice their proactivity and empathic listening skills. The project was floundering but we soon devised a plan to implement an environment mostly comprised of simple software agents who had personal knowledge of various events, and who would discuss things if the right questions were asked. The technology was little more than "agents as Ask Networks" with a few additional wrinkles (mostly conversational dependencies).
By the summer of 1996 this project was near completion and soon due to be shelved.