About SoLOrganizational LearningPublications & ResourcesConsulting & ProgramsCommunities & ConsortiaJoin SoLMember Access
Home > pra (OLD) > epsilon.html
The Learning Initiative at AutoCo

Origins of Epsilon's Learning Project

AutoCo's interest in systems thinking (which later included organizational learning) began in 1989, when Peter Senge (who was developing a Center for Organizational Learning based at MIT) and Russell Ackoff (professor emeritus at the University of Pennsylvania's Wharton School) started giving monthly presentations in AutoCo's Executive Development seminars. A manager in AutoCo University [AU Manager], who sponsored the monthly training sessions, was interested in testing these systems thinking concepts in one or more live business settings at AutoCo.

Note how seeds were sown for some time before a project opportunity emerged at AutoCo. The company "pulled" the effort in, rather than being "sold" a bill of goods.

As at other large companies, AutoCo internal "change agents" had to consciously decide between a "bottom-up" or "top-down" approach in any given initiative. From the beginning, this effort took the "bottom-up" approach. This meant it would be easier to implement, but harder to expand to fit the larger AutoCo system.

Note how this very concrete learning effort begins with three abstract "governing ideas." It's not clear to outsiders whether a concept like "thinking differently" meant the same thing to different AutoCo managers and executives.

AU Manager: We had just finished our first executive education program for the top 2000 people worldwide. It was a gathering from the four corners of the world, and it was quite a happening. The question was, what should be in the second round?

We decided on a "bubble-up" rather than "top-down" model. We went around the world and interviewed executives and asked them what was on their minds. What would be of the greatest interest to them? Three issues surfaced: globalization, thinking differently and leadership.

Underlying the first two issues was the pervasive issue of change. We then went about exploring what would be a senior executive program built around these three themes. There was controversy in presenting them to the top of the house because the top felt they might not be ready to get into all this subject matter. Nevertheless, they said, "press forward."

AutoCo called upon well known academics and consultants as part of an education program for senior managers. Each issue had its own leading "expert," with the exception of systems thinking, which had two strong voices: Peter Senge and Russell Ackoff.

These two experts held, in common, the view that an organization's work could not be understood in fragments. AutoCo's managers responded to this message because it helped with the perennial problem of miscommunication and conflict between functions. Intellectually thought provoking and personally compelling, but very abstract, the ideas in systems thinking were flavored by their academic origins. AutoCo managers were intrigued but they challenged the academic stance by asking, "How are these abstract concepts applied?"

In the arena of "thinking differently," we came across two outstanding voices: Peter Senge and Russ Ackoff. The more we dug into the area, the more we found a very significant message coming out of Senge's and Ackoff's world views.

By late 1989, early 1990, we had both Senge and Ackoff doing a program at our center every other week. In our analysis of participants' reactions, there was a large voice saying, "The ideas are intriguing, but I don't see how I would play them out in my ballpark -- on Monday morning!"

In all fairness, we had asked both Senge and Ackoff to take us on a broad journey, and not to focus specifically on application. It's not surprising that participants were intrigued with it, and saw its depth, but they were right in feeling that there needed to be an ability to see further down that chain of, "What happens next?"
The AU Manager took the response of senior executives as a challenge to find a way to operationalize systems thinking throughout AutoCo.




The comments people asked about application led to their being critically challenged.




Is it possible for executives who have spent their lives thinking in a particular way to change their thinking?

That's when I formulated the challenge for myself and AutoCo University to continue to pursue this subject area. I had the good fortune to be able to sit in on numerous sessions with Peter and Russ, and in the fifth or sixth session, the ball bearings started to rotate in unison.

I asked Ackoff publicly, "Russ, I've been sitting here for several sessions, it's an outstanding message, but I'm still having trouble digesting it and its implications."

Russ turned to me and said, "Well, that's because you'll never get it."

I turned beet red. Here I was standing in front of 50 executives, and the room was dead silent. Then Russ let me off the hook. He turned to the group, and said, "And you won't get it, either. We have built up over 400 years of methodology of 'reductionist' thinking. It is so powerful, so pervasive, that probably your children and their children will have a much easier time. For you folks, it's going to be tough grind."

Down to my socks, I understood that you can say you understand it, and still not understand it. The implications were absolutely profound. Organizational learning didn't mean letting go of analytical processes. It meant complementing and supplementing them with synthesis or systems thinking. It's not such a clean thing -- "Just throw out all the traditional tools, my past life -- and switch into new formulas." It means learning something in addition: the "and," not the "or."

AutoCo's introduction to systems thinking thus represented a challenge: How could the organization make use of systems thinking in a business context?

In 1991 a diverse group of AutoCo managers attended a series of five two-day training sessions on the core competencies of a learning organization run by the founding staff from the Center for Organizational Learning. To the surprise of the AU manager, given the abstract nature of the materials, the first audience exposed to the concepts was receptive. This audience was composed, in part, of managers from product development. In particular, the Epsilon Program Launch Manager expressed keen interest. He was responsible for the vehicle development program to design and build the next model of the Epsilon luxury car.

Past experiences in developing cars were powerful influences in trying to create a better approach on the Epsilon program.




Although good quality and timely delivery are held up as critical, achieving them does not make cost overruns acceptable.




This is not the only example where attributing good performance to team dynamics was not accepted as a convincing explanation by executives. It seemed that executives expected and listened for technical innovation, rather than management changes.

Do the ways in which success is achieved in large corporations have to have task oriented, replicable type answers to be accepted?

Launch manager: I had worked on the Delta program [another vehicle program] for several years as the Business Planner and Launch Manager. We had discovered, a year before Job One, that the program was 17 months behind schedule. So we quickly organized a 100-person launch team and we put the program back on schedule, with quality that was better than the first car. We met all of our program objectives except cost, which we knew from the beginning we could not meet.

I remember a meeting where a Vice President listened to us present the reasons for our success [on the Delta program]: team leadership, and the fact that everyone had the same goals and knew that they depended on each other. "That sounds really great," he said, "but what did you do?" He finally said it must have been a fluke, and that was the end of it.

There was no learning from the experience. It bothered me a great deal. AutoCo is in love with managing by crisis; without a crisis, we don't know what to do. I resolved to learn how to produce a car launch without a crisis.

In July 1991, the AU Manager, along with an internal consultant working with the Launch Manager, wrote a letter to Peter Senge, director of the MIT Center for Organizational Learning. They requested an active relationship between MIT researchers and the Epsilon Program.

Project Engagement Clinic

The formal partnership between the two organizations started with a project "engagement" clinic in September, 1991. The goal of this clinic was to engage one another in asking difficult questions about readiness for learning, and explore possibilities for a partnership which sought business improvements and research results. Attending the meeting were the key people in the project -- the Program Manager, Launch Manager, and Body Engineering Manager for the Epsilon Program, two internal consultants on process improvement (who were to integrate systems thinking at AutoCo) and five researchers associated with the MIT Center for Organizational Learning. One researcher had visited AutoCo and conducted interviews in advance.

These interviews were summarized in a report that singled out several key issues:

  • The Launch Manager wanted the MIT project to focus on improvements in the Evaluation Prototype [EP]. If the EP could be made to "work" as it should, the car would be successful, he said. He wanted to create a climate within the team which reinforced more effective cross-functional communication, more responsibility for objectives, and less "games-playing."

  • The Epsilon Program Manager (the Launch Manager's boss, with overall responsibility for the Epsilon project) believed that concentrating on one prototype, like the EP, was too narrow. He felt it was necessary to look at AutoCo's product development paradigm. Present product development management practices controlled resources ineffectively, treated suppliers with indifference or hostility, and had a history of not achieving quality, cost and time objectives. The Program Manager mentioned the inability of a program to reach a point where management could say, "Enough changes this time."

  • AutoCo's vehicle program management needed to see tangible improvement. The new product development program officially involved a 48-month time period, which was compressed to 42 months, and had very ambitious deadlines. Chronic dependency on "heroic" efforts took key people away from the planned product development process. Some people felt certain that it was just a question of time before a real crisis became obvious.

  • MIT's researchers wanted the project to further the investigations being done in the Center for Organizational Learning. Supporting research would require more than a relationship where MIT played the role of expert consultant and AutoCo took the role of client. Instead, both organizations would have to collaborate in a systemic approach to improving the product development management practices and understanding the results that were achieved. Measurement would be important in determining if improvements occurred over time.

Following the project engagement clinic, several meetings were held to determine the project focus and how MIT researchers would work with the Epsilon team. In January of 1992, a core "learning team" of managers from the Epsilon program began meeting regularly, at one- or two-month intervals. The core "learning team" consisted of ten Epsilon managers (including the Program Manager, Launch Manager, Assembly Manager, Finance Manager, Body Engineering Manager, and Purchasing Manager, and two internal consultants). The lead MIT Researcher was present as a facilitator. These meetings provided opportunities to plan, learn techniques and tools that facilitated learning, and practice using those techniques between meetings.


Sidebar: What are "learning laboratories"?
"Learning laboratory" is a term used by the MIT Center for Organizational Learning for a workshop, often also called a "managerial practice field," where people come to develop new skills, cycling back and forth between study and practice.

The learning lab is different from a training session in that over time its concepts and values are intended to become part of, to be integrated into, work issues and job settings.

In sports, the arts, and the military, teams are accustomed to practice sessions, simulating real events where members learn from mistakes and each others' examples. Similarly, learning labs simulate normal business settings, providing an opportunity for management teams to practice together and make mistakes without penalty or the pressures of performance. Participants learn new tools by applying them to the issues they face in their day-to-day jobs. The MIT-COL learning labs at AutoCo focused on skills of conversation, reflection, and systems thinking.

The two-day learning labs at AutoCo were designed and facilitated by Dan Kim from the Center for Organizational Learning and selected managers from the Epsilon program team. The Launch Manager was the first AutoCo facilitator, and he was directly involved in all the learning labs.

The designs were based on the experiences of the first core team, and modified after each session in light of the experience of new labs, the makeup of participants, and new issues faced by the program. Some participants in early labs became facilitators of later labs. Four learning labs were run during the program. Over 100 people -- more than a third of the full-time, dedicated engineers on the program -- attended.

The learning labs at AutoCo alternated between conceptual sessions for learning new tools (of conversation and systemic thinking) and exercises for practicing their use. These exercises were deliberately designed so that people could consider their own work issues with perspective that came from the deliberate telescoping of time and space. For example, a computer simulation allowed participants to spend an afternoon working together through a product development process that would normally have required three to four years.

For more about the learning labs at Epsilon, see Theme 3: Learning labs: Teaching techniques for thinking differently. For theoretical information on designing managerial practice fields and learning labs see Kim and Senge, "Putting Systems Thinking into Practice" in System Dynamics Review, 1994, Vol. 10, Nos. 2-3, pp. 277-290.

After meeting for eight months, the core learning team designed a series of learning laboratories, through which engineers on the program were to learn and apply techniques of systems thinking, working with mental models, and team learning. Those learning laboratories took place in September of 1992 and February, May and August of 1993. The efforts of the Epsilon program to apply organizational learning concepts spanned the following three years.

[Back to the Abbreviated Table of Contents]