How would you define the "new engineer"?
The "new engineer" is an engineer who is capable of functioning well in the world of the 21st century, where telecommunications will vastly shrink distance and time, making all business international. I expect major shifts in the role of the engineer, from being primarily a problem solver to being an good listener and effective facilitator of client projects. The new engineer will not only have to satisfy all the engineering requirements, but in effect, function as the client's partner in the many domains that use computing.
What modifications will GMU need to make in order to teach the new engineer?
The changes need to occur on many fronts. One is the curriculum front. People often refer to the curriculum as the content of the engineer. Changing the content of engineering is not an exercise that can or should be accomplished in isolation. The considerations of industry, students, and faculty must be taken into account, this is not always an easy or effective task. I personally do not like to view the curriculum as content alone. I prefer to teach the habits of a profession as well as the technical content. The habitual routines that people use to solve a problem and assess a situation will be different and will change as technologies develop. The practices which will be valuable in the years ahead are different than the ones than the ones we are used to. For example, being an adept user of the Internet is a relatively new practice which is important for engineers to master.
One of the important aspects of all this is that we must constantly be learning. Our engineering tradition tells us that we ought to be building a curriculum that will prepare a student for a lifelong career. But this is no longer the case today. Careers don't last a lifetime, technology is changing, practices are changing, and we cannot expect that the education you get in the 1990's is what you will need in 2010.
Continuing to learn must be the practice of an engineer. We need to work on that. Traditional canned classroom materials that make students into passive receivers of engineering knowledge do not give them the practice of learning what they need to know. The Internet provides tools that they can use to find the knowledge that they don't have and need to have in order to work on their projects. We need to encourage them to find these tools and how to use them well.
Are you stressing continuing education or professional development?
No, we are stressing a change in habits. We need to understand what we need to learn to get the job done, and who are the experts we need to consult with.
In my observation many students fall into a bad mood when presented with a challenge. For instance, if given a homework set that isn't clearly spelled out, they get upset, because they don't know how to grapple with the ambiguities. In some ways the tradition of presenting students with well defined problems is exactly the opposite of what happens in the real world. In the real world a problem may not be completely defined. When a boss specifies a project they may know what the expected result is but they usually don't know how to do it.
How effective are the educational tools that you are developing for the Internet? Are you competing with or complementing conventional education?
You're pointing at what appears to me to be a deeper issue, that is, the question of how much you can put into the box. I was discussing the future of the university with another person, and speculating as to whether 10 or 15 years from now there will be so many educational materials on CD ROM that will run on your personal computer at home, that you can give Jane these modules to work on from 9am to 3pm at her home. After you do this day after day, would you say that she is educated? Most people that you propose this to don't think that this would achieve an education.
There are a lot of things that you learn that you cannot first put into a box. There are many things that you learn from interacting with other people. A box will tell you about the syntax of programming languages, that will let you write programs, submit them and grade them, because programming is a fairly sharp, well-specified area. These programs are typically only one to two pages long. Such simple procedures can be taught by interacting with the box, learning how to deal with the syntax of a program, how to test it, redo it and resubmit it automatically. Not much will be lost. If the student is stuck, then a teacher might be consulted. However, more sophisticated jobs such as when a student team works on a design project for an outside client, solve the problem, demonstrate their resourcefulness, and satisfy the client is not something that can be done with a box. You can only do this by being involved with the client and the team. Learning how to make your engineering knowledge available, in the context that they are operating. You cannot teach this in the classroom. It can only be taught by putting students in a specific situation.
Once I was reading an essay that intrigued me where somebody was talking about all of the things that are possible with computer tools for education and what we have certainly experimented a lot with in the past is boxes that present material to you. There is a lot of talk about another kind of box, an assessment box, which will test you. So box 1 is a box that teaches you material - call this the teacher box, then you go to box 2 that tests what you have learned - the assessment box. Then you can have feedback between the two - if you are not doing well as measured by the assessment box, it will ask the teacher box leads you a bit further. Now the "you" in this conversation is the student. So we are willing to accept the possible existence of a teacher box and also the possible existence of a tester box. Well, while we're at it, why don't we ask, what about the possible existence of a student box? This is a third box that sits there and gets taught and learns until the tester is satisfied. Now we have three machines interacting with one another. There is something deeply unsatisfying with this scenario. It is the loss of involvement. You learn things by being involved with the community. No box can teach you how to involve yourself with clients, teams, etc. For example, how do you teach trust? Some things must be accepted on trust. There are a lot of things you trust, that you can only learn by being involved with other people. This scenario of high tech education, where you have high tech box for teaching and another high tech box for assessing leaves open the possibility of having a third box that replaces the student. So you put all of the three machines in the closet and say, now, lets have a class! Machines are useful as tools, as being depositories of information, for organizing and presenting information, but in the end they are only tools.
So how can you assist teachers, especially in K-12, to leverage their resources with these high tech tools?
The CLIN component is trying to help implement the regional consortium which will be the prototype for other regional consortiums around the country. Just approved is a plan that would deploy eight CLIN sites into the schools. (in Arlington, Alexandria, Fairfax, etc.). Each site would come with video, internet access, and oriented towards making these things available on affordable computers.
Fairfax Chapel Square will be a training center for the teachers. Initially Chapel Square will be one of the nodes and then they will take over the training and we will help them. One of the reasons for creating the regional consortium is our deeply held belief that education not a local activity, but a community or regional activity. We need to teach in consideration with the needs of the community. This will also benefit GMU because we will have more and better prepared students going into science and engineering.
What in your background has influenced you to form the Center? What leads you to hold the principles that the center embodies?
Primarily, a long held interest in the question of curriculum. This dates back my days as an Assistant Professor at Princeton around 1970. One of the questions on the table at that time was whether the core of computer science should include operating systems. At that time only the mathematical subjects were given core status -- computability theory, switching theory, parsing and automata theory, algorithms analysis, and numerical analysis. Compilers, linkers, loaders, operating systems, graphics, programming languages, robots, expert systems, and databases were considered applications, not core material.
I was convinced that operating systems ought to be part of the core and made the case in 1971 in an NRC report and major article in Computing Surveys. I was convinced that operating systems were based on sound scientific principles that formed the basis of a theory. With Ed Coffman I wrote the book Operating Systems Theory in 1973 that confirmed this stand for all time. My next major curriculum-related activity was in 1980, when I was ACM President. I observed an alarming outflow of systems-oriented people from universities to industry. I thought that if this trend were to continue, it would harm the mission of computer science departments to the detriment of students. I was heavily involved in making the case for universities to revise the salary scales for CS faculty. We were successful.
I went to NASA in 1983, and for the next 8 years got to observe an outsider's view of computer scientists. In the mid 1980s I led an ACM/IEEE task force that produced a report about the core of computing, which laid groundwork for new approaches to structuring curricula. In 1991, I returned to the academic life, having been goaded by the rising criticisms of universities. I came to George Mason because I concluded that the entrepreneurial spirit of the faculty would allow experiments in the way we decide what to study and how to study it, leading eventually to reforms that would blunt or eliminate the criticisms. I soon discovered that people at ARPA had similar concerns in the specific domain of HPC (high performance computing) and with Daniel Menascé submitted a proposal that was successful. We started a number of related and interlocked projects to support the ARPA work, and in order to give them organizational coherence, we founded the Center for the New Engineer (CNE).
What is the goal of the Center?
One of our premises is that successful research must be connected to the undergraduate curriculum, and that a successful department must be an active member of the regional community. The goal of the Center is to show how to establish a feedback path from research to curriculum, both undergraduate and pre-college. This means to establish a social system, supported by high performance computing and communication technologies, that values the transfer of new knowledge about advanced computing technologies from researchers to students at all levels. What is being learned from HPC research must not be restricted to the immediate circle of researchers and their graduate students. In my view, this is the key to building a very good, high-quality computer science program at GMU. Moreover, our involvement with the regional K-12 schools can have a significant positive effect on students applying for admission as engineers at GMU.
What would you like the Center to achieve five years from now?
Our main interest is integrating what we are learning from research into the curriculum. It appears to the outsider that we run a very expensive research enterprise that is not benefiting most students. The graduate students do benefit, but the undergraduate students seldom come in contact with the research. Some people have said that the young student learns more about advanced computing technologies from the Discovery Channel than from their own computer science classes. We would like to change this. We would like students to be learning advanced technologies before they get to the Discovery Channel. We would like to turn the tables, so that we would be able to hear this conversation: A parent watching Discovery Channel exclaims: "Hey look at what they are doing with that supercomputer. They are splicing genes." To which the student responds, "Oh, we did that in computing lab. No big deal." We are not expecting undergraduates to be expert users. We do want them come into contact with the new technologies, to know that they exist, and to know of their capabilities.
When we talk about the feedback, we must not be too limiting in our interpretation of what this is. It must not be restricted to university research, but must include the research of regional businesses, who are also inventing new knowledge. It must also take into account the expectations of organizations that hire our graduates, about what competences those graduates must have.
So part of the vision is to develop these feedback paths, to continue to experiment and to make these resources available on the Internet for other academic departments within GMU as well as computer science departments, and even available at the high schools?
Yes.
Would you comment on the researchers working in the Center?
We interviewed many candidates, made sure that they had some of the expertise in the technologies we are using, and were interested in forwarding the mission of the Center. We were very pleased (with their progress) in the last few months. They are starting to teach us things we didn't know about, finding information from the Internet, coming to us with new gadgets that solve some problem that we just encountered a week ago. I'm very pleased that that's happening, The team has come together in about three months, is beginning to get their own identity, is learning about new tools and bringing them to our attention, getting plugged into the national networks, which is exactly what has to happen.
The Center has a lot of ties to other institutions, can you talk about the faculty participation as well as the participation of industries?
The faculty are the experts on advanced computing technologies. Those who participate are taking on a responsibility to make some of their expertise available for undergraduate students. Some will do this by preparing interactive texts through "authorware" tools. Others will do it by preparing demonstrations of software systems used in the research domain. Still others will do it by creating workbenches that permit students to perform experiments with the advanced technology. There are many open questions. How do we make advanced technologies accessible to the young mind? How do we design homework or lab projects that will use the laboratory materials? How can we design industry participation? One possible form of industry participation that is being greeted warmly by industry representatives is to propose projects for students of the new Design Exhibition course (being piloted by Dr. Menascé) and to provide staff time to mentor students.
Isn't there another level of industry participation, such as with Lotus and Action Technology, where they are providing an enabling technology?
Yes, there is a working partnership with Lotus to experiment with Notes in the building of a "collaborative space" in which instructors and students can organize their collaborations over the net. We are a member of the Lotus Education Consortium, a group of universities who are conducting experiments of this kind. We are also a business partner with Action Technologies; we will test their workflow management software and use it to help organize the administrative processes of courses and of the department. Through our subcontract, we work closely with CLIN (Community Information and Learning Network), who are very good at finding affordable ways to bring networking and educational software to remote sites.
So it seems like the Center has a wealth of collaborators in the community and in the university.
Indeed, we have many ties with others of similar interest in the University. Professor Chris Dede of the Graduate School of Education is a co-investigator on some of our projects. He is helping us to assess what the students are learning in the new approaches we are testing. I mentioned earlier that the curriculum is focusing on the questions, What to study? How to study it? Well, assessment is focusing on the question, How do we know when they know it? We are also working with the Instructional Development Office (IDO), co-directed by John O'Connor and Randy Gable. They are dedicated to helping GMU faculty use technology for more effective instruction. We are working closely with faculty from the Computational Sciences and Informatics (CSI) Institute, who have knowledge of many HPC results that our students ought to learn. We are working with other departments in SITE in the same way.
What is "Sense 21"?
That stands for "a new engineering common sense for the 21st century." I taught an experimental course under this designation in the spring of 1993 and am repeating it this spring. My intention is to make students observers of the existing engineering common sense, to speculate about the new common sense they must have to be effective in the 21st century, and to position them as early practitioners of the new sense. The alumni of the spring 1993 course formed into a club we call Sense 21 and meet once a month (over pizza) to discuss new books that appear to offer powerful new interpretations, thereby to hone our skills at observing and making our own new interpretations of what is going on. When the current group completes the course, they will join the alumni group. Some day, I hope that there will be enough interest in this course that it would be required of all our majors.
Interview Concluded