Letter to the Editor - Should science and engineering undergraduate education be merged?

Author:  Sravisht Iyer
Institution:  Johns Hopkins University
Date:  April 2005

The definition of a scientist and engineer can be as varied the people who define them. Engineers are generally considered to be those who create products to meet human needs, while scientists are those who devise instruments and experiments to test their theories. A more limited definition might describe engineers as those who make the instruments that scientists use to take their measurements. This point of view, however, ignores the fact that engineers can only solve complex problems after developing a keen understanding of the underlying scientific phenomena, or that scientists can only convert their experimentally acquired knowledge into working systems after understanding the engineering required to scale up the process. So if engineers ultimately need to understand science in order to solve problems, and if scientists must ultimately embrace engineering in order to create useful products or processes from their discoveries, why do the two groups often have such a difficult time working together?

One answer suggests that this conflict can be traced to their education. Historically, engineers and scientists are separated during the education process. At the University of Hawaii, for example, there is the College of Engineering and the College of Natural Sciences. The two colleges are not only separated administratively, but geographically as well. The separation is magnified by the fact that humans are essentially creatures of habit that align themselves with likeminded colleagues. Consequently scientists and engineers generally pass the majority of their careers relatively segregated from their engineering/science counterparts. And when they do work together, the segregation often persists in the separation of project tasks along "engineering" or "science" tracts. It's only natural for a scientist or engineering to gravitate towards their field of study and to believe their approach will yield the best solution. It's within their comfort zone.

Part of the reason for this is that science and engineering undergraduate programs train their students to think differently. Engineering is considered a professional program, one with governing bodies that accredit undergraduate programs using relatively standardized and rigorous guidelines. There is a good reason for this. Engineers construct systems, manufacture plants, and other civil infrastructure that are used on large scale by large numbers of people on a regular basis. From this perspective cost reduction and avoidance of liability become important outcomes. This necessarily translates into courses that emphasize problem solving with closed solutions that stress over design, elimination of uncertainty and variability, and perhaps more importantly, avoidance of liability. Certainly some engineers work with scientists on problems that present new challenges needing new solutions but by and large engineers are trained to solve problems and to design processes that eliminate variability and to generate consistent and predictable results.

Scientists, by contrast, are trained to ask questions, or pose hypotheses, that lead to clever experiments that yield answers that generate more questions. A scientist is happy when an experiment generates a hundred new questions, all of which must be answered before the system is really understood. In science there is really never an end to the number of questions that can be asked, or the experiments that can be run, before there is enough knowledge to truly understand a system. This scenario, however, is an engineer's nightmare. I can't remember how many times I've heard engineers complain that if they considered everything the scientists wanted them to, they'd never solve the problem. This is true, to some extent, because engineering solutions require that those variables that do not directly affect the solution be eliminated while those that do be strictly controlled.

Let's consider an example that applies to Hawaii. The island of Oahu has an imposing mountain ridge that runs the length of the island. Splitting off this main ridge are numerous side ridges that slope down towards the shore line. These side ridges create a number of three-sided valleys. As it rains almost daily along the top of the main ridge, most of these valleys have a river or stream that carries the rain water that trickles down the sides of the ridge walls, eventually emptying into the ocean. Every once in a while periods of great rain fall can occur, as happened in the Manoa valley late last year, generating floods. This becomes a problem as people move into valleys and build homes. As on the mainland, organizations such as the US Army Corps of Engineers have built floodwalls and diversion channels to contain the flood waters in order to minimize flooding. The best engineering solution will construct diversion channels that are sufficiently large to handle the "every once in one hundred year" flood. To avoid filling and blocking of the channel from deadwood and other natural debris, the solution will also want to cement large lengths of the channels.

So where is the conflict with science in what appears to be a perfectly engineered solution to a real problem? As it turns out the earthen stream beds contain rocks and other materials that support the growth of microbial biofilms. The bacteria in these biofilms are extremely important for the uptake of nitrogen and phosphate. Unfortunately, the cemented diversion channels lack the required shape and organic matter to support the same extensive network of biofilms. Consequently, much of the nitrogen and phosphate makes its way to the open ocean. As the immediate coast off Oahu possesses shallow reefs with low mixing, the nutrients end up providing food for plankton and other small free floating organisms. Their increased growth increases turbidity, which not only decreases the clarity of the beautiful Hawaiian waters, but also induces algal blooms which damage coral reefs.

In the above example, the engineer solved the problem by only considering those variables that were important to achieving a singular objective: the diversion of flood water in the worst case scenario. The best engineering solution excluded any consideration of the potential impact on the local environment, both in the short-term (by cementing over a natural stream bed), and in the long-term (by damaging an ecosystem of microbial biofilms). While this example was relatively simple and applied, it should also be kept in mind that advances in nanotechnology and molecular biology are beginning to erode the traditional academic borders between engineering and science. This is leading to a host of new problems that can only be solved if they are approached from an interdisciplinary perspective. Indeed, this has been a theme pulsing through NSF for some time now.

Consequently, there is an emerging need for engineers to be able to solve open ended problems with no set solutions and for scientists to be able to focus their research outcomes on more tangible objectives. Some engineering education circles have proposed the incorporation of interdisciplinary research. This is a fine goal but how do you do it? The point here is that engineers can't teach science and vice versa. One can create multidisciplinary teams of faculty but this is difficult to practically administer, particularly across colleges. A more promising approach is to place the engineering student directly into the research laboratory of a scientist and vice versa for the science student. It would be great step forward if undergraduate science and engineering programs developed programs that required their students to pass an extended period of time performing research in the labs of their science or engineering counterparts.

In the Sea Grant Marine Science Undergraduate Research Fellowship (MSURF) we have applied this principle. The program recruits students from all engineering and science disciplines and endeavors to place the students in projects that are recruited from any science or engineering discipline. The only common link is that all projects must be concerned with marine/ocean science or marine bioproducts engineering. To our surprise we have had projects proposed by faculty from electrical engineering to zoology. In this edition there are mechanical and chemical engineer students who present papers on research conducted in the labs of scientists. While I've found some faculty to remain cautious and to hesitate at accepting a student from unrelated disciplines, it has also been my experience that faculty who have accepted students outside their discipline have been pleasantly surprised at both the ability of the students to learn and perform quality research, but also in their ability to bring unique perspectives. Likewise, the students have left with a renewed confidence that their career is not limited to their own discipline, and also with a greater sense of how the principles of their discipline can apply to another field.

To some extent I believe engineering and science education should both be kept separate but also merged. It is important to not loose sight of the importance of undergraduate education in a traditional science or engineering major. The compromise, in my opinion, is for students to pass a significant period of time conducting research projects (i.e. not just courses) in a research laboratory of an entirely different discipline. Preferably, engineers should work in the labs of scientists and vice versa. I believe this process to be the most effective and in MSURF the data to prove it. But most importantly, I think it is the responsibility of faculty to support this process.