This is an essay, rather naive, written while I was in college. You can observe the excessive use of "we" to sound more authoritative, in fact I had to get a co-author for this express purpose!
Mathematics obviously forms an important part of engineering education, since it provides us with the tools to solve engineering problems. We were never good at mathematics, but still we would like to describe our experiences of learning engineering mathematics. Perhaps it would be enlightening for us and for others to discuss such learning experiences with fellow students.
Engineering studies is unlike school education. In School math is taught as a subject for its own sake. In an Engineering course mathematics serves as the foundation on which engineering disciplines are based, so it is not here on its own right, but only to help understand the other engineering subjects. What we have seen is that the engineering mathematics syllabus of Kerala University emphasizes an abstract approach; an approach that does not makes use of physical reasoning, and keeps the mathematics isolated from the engineering subjects. The mathematics problems always deal with x and y, and never with heat and pressure or any other things which may be encountered in an engineering subject. That is, the maths course just deals with some problems and some methods to solve those problems, it does not give any indication of how to apply those methods in to a real world problem.
Take the case of Differential Equations. Differential equations are of great use in most engineering disciplines, but our engineering math course, deals with it in a way that is completely boring. The syllabus just required the students to learn many methods to solve the equations. But, before the equations can be solved, the physical problem has to be modelled and the differential equation formulated. The syllabus completely ignored the modelling aspect. Perhaps the most fun part of solving a differential equation is forming the differential equation , but we missed that entirely.
Similarly vector calculus was taught in a way that was very abstract. Physical reasoning was not used much in explaining the methods of vector calculus. Using physical insights from electrodynamics would have made it more interesting to learn and to apply, instead the syllabus just required to learn many formulas for div, curl and we didn’t have to bother what these all meant, we only had to learn how to solve the usual type of problems that came for the university exams. The situation is the same in most of other topics as well. The magic of Fourier series, that once you had learned it you get the amazing ability to write almost any function as the sum of a series of sines and cosines, was lost on us, we just saw it as some more difficult stuff to be learned to get pass in the exams.
Another issue with our engineering mathematics is that it completely ignores computer based techniques. Mathematicians may prefer to work with pencil and paper, but using computers to solve problems is very prevalent in many engineering fields. The engineering mathematics is so focused on ‘pencil-and-paper’ approach, that it ignores many interesting methods of solving problems using computers.
The problem with these paper and pencil methods is that, some of them are quite easy to do using paper and pencil for small problem sizes, but the same methods are cumbersome when implemented on computers. We are all taught to evaluate determinants. The most commonly used method is the Laplace expansion method. It seems to be the easiest method to use to evaluate determinants by hand, and we rarely have to solve determinants greater than 3x3 in class or in exams.
But, problem occurs when we have to implement a program or function to compute the determinant of a matrix on a computer. The Laplace expansion method is a very unsuited for implementing on a computer because its running time is order of the factorial of n, O(n!): for a matrix of order n, it will take a number of operations that is proportional to the factorial of n for computing the determinant using the Laplace expansion method. That is to compute the determinant of a 20x20 matrix, at the rate of one operation per nanosecond would take 20! * 10^-9 seconds. It will take several years to finish.
Of course other methods of evaluating determinants are taught too. The Gauss elimination method, converts a matrix in to triangular form by using elementary transformations, once the matrix is triangular, its determinant can be taken by taking the products of diagonal elements. This method takes only O(n^3) operations, that is it can be done by n^3 number of operations, where n is the order of the matrix. But when we are calculating on paper by hand the Laplace expansion feels much easier than the Gauss elimination, and when we are doing a computer implementation also, we are tempted to use the Laplace expansion, considering the ease of doing it by hand for small matrices. But this factor of computational efficiency was never mentioned, because the syllabus didn't care about computer implementations at all.
An approach that uses physical reasoning cannot be considered rigorous, and pure mathematics never makes use of physical reasoning. But, engineering mathematics should be closely related to the engineering disciplines; perhaps that will make it more fun to learn.
May be we were not interested enough in mathematics, may be our text books were not good enough , may be we did not listen to class carefully, but in the end we did not have as much fun with engineering math as we could possibly have had.
Ajoy W
Kiran J.L





