During a recent meeting with engineering-school faculty and alumni, we talked about whether their college should educate generalists or specialists. One of the graduates explained how his broad education let him solve a problem with fundamental information that bridged several specialties. One of the engineers with a deep knowledge in a narrow area countered that today many companies need engineers with specialized knowledge so they can "jump into" a problem right away without a "warm-up" period. I can see both sides of the generalist vs. specialist debate.
In electrical engineering, undergraduates often specialize a bit, perhaps taking more analog than digital electronics courses. But they receive a BS degree with a good understanding of many facets of electronics. In graduate school they can continue their education in narrower fields. Undergraduate engineering programs educate people about how to approach and solve problems, and how to think critically and examine problems from several perspectives.
The general knowledge instilled during four years of college also helps graduates evaluate a field and determine whether they want to continue in it. I know science and engineering graduates who have become surgeons, physicians, teachers, entrepreneurs, patent attorneys, and so on. The generalist approach served them well. This approach also lets people who aim for more education benefit from a variety of experiences in their discipline. So I would not recommend trying to push undergrad engineering students to become specialists in four years.
On the other hand, when companies and universities advertise job openings, they usually have a long list of specialized requirements. I found this example of job requirements on the Internet:
Minimum five years of embedded FPGA/ASIC design and/or verification experience;
Three-plus years of experience using System Verilog;
Strong working knowledge of OOP verification and verification environment;
Experience with OVM/UVM verification methodology;
Good verbal and written communication skills;
Self-starter who can work with minimal supervision in a team environment on site;
Experience with scripting languages (e.g. Perl, TCL).
Generalists need not apply. So here's my advice: Go ahead and specialize as you see fit either through an advanced degree or on-the-job training. But keep an eye on general knowledge in your chosen and related fields. If you want to specialize in motor control, for example, you should know how to write code in C, simulate control algorithms in MATLAB and Simulink, use LabVIEW, and so on. It also helps to know how to go to the shop and quickly machine a motor coupling you need to test a motor. You might become a specialist with a generalist's knowledge of many things, or a generalist with pockets of deep knowledge in a few areas. We have room for both types of engineers.
Readers, what do you think? Tell us in the comments section below.
ttemple, interesting observation. Another aspect of the debate, in addition to generalist vs specialist, is the amount of theoretical vs practical training should be required. Some time ago, it was deciced that many engineers did not have a good enough grasp of the science and mathematics behind their fields. As a consequence, many programs added more science and math classes. I guess it is not enough.
William, the situation you mention with the FPGA is an interesting one. Frankly, if you have a job that is quick turn around for a parrticular part what you need is a consultant, not an employee. Frankly, the idea that you need years of experience with something that specific is a bit of a problem. In the current technology environment there is really not very much that is the same for that long. Witness the promise by chip manufacturers that a particular part will be available for ten years. This is for areas such as automotive, where a model may be built for that long and the manufacturer does not want to have to certify a new chip during that time. Let's just say that the guy who has five years working with one part is not that likely to innovate (I have lots more examples).
I have run into the situatioin similar to you mention where headhunters want an entry level person with six years experience. It was a while ago, but it was not unusual. The request was for a consultant with five years experience in a particular operating system that had only been out for a year or so. It was a standard requirement. Most headhunters have no idea of what they are recruiting for. They are only there to take the load off the hiring manager.
I agree with you that what is useful is a grasp of general engineering principles and techniqies. Many projects do not fit neatly with one area or another. I know several people in the aerospace and robotics fields who, while starting with one degree (mechanical or electrical) get a master's in both fields. In the FPGA example, I recently ran into one organization where FPGAs were used extensively. They were programmed by both electrical engineers and software engineers. It seemed like whoever was available at the time.
Your comment on starting in Physics is interesting. An observation that I have made over the years is that in general engineers don't seem to understand physics well enough. Or if they do know physics, they doesn't seem to BELIEVE physics.
I think you're right on target, AGHock. We need both types of indivduals. For some people, it definitely makes sense to specialize, especially if they've found their technical passion. But many engineers change jobs and careers more than once and may not always use a specialization. Still others end up filling jobs outside of engineering, where their technical accumen may be useful as background only.
There are certainly places where both types are needed and will fit well. A very large organization may need somebody who is a master with some particular model number FPGA, and nobody without a lot of experience with that part will do them much good. On the other side of the picture, how many places woulkd that person fit? and how soon would all that expertise become outdated and of marginal value? Besides that, how possibly can a university provide such a specialization that is not "past it's prime" the day the student graduates. And what if that same specialist needs to add an interface and amplifier for an A/D converter?
Headhunters seem to be a group that is either seeking an "entry level person with 6 years experience", and sometimes that experience on a device that has only been in production one year, or they want expertise in a wide range of areas, well beyond what is reasonable.
But a good grasp of a general engineering should be useful in a number of areas, and the background is vital for being able to develope a specialization.
I don't design logic chips, I design custom automation, among other things, and it is not economical to design a chip that would only be used in a dozen systems. So the ability to work in a more generalized manner has been quite worthwhile for me.
BOTH Generalists and Specialists. They do not have to be exclusive in an educational institutiion, though curicula tend to promote one or the other. This does NOT have to be.
Further, many of us do not get a good handle on what we are good at until some years into our career. I think that what is really needed is to be able to learn more, academically, about spicific areas without having to devote full-time to a class program. In fact, many institutions basically force you to enroll in a degree program to get those graduale level courses. That should NOT be so.
Writing as a research support factuly person at a mid-level engineering univerity, with lots of industriial experience.
A skyscraper without sufficient foundations will fall down.
While the Leaning Tower of Piza is an interesting icon, it is necessary for graduates to have a sufficient grounding (foundations) in their subject so that they don't get blown over by the winds of change.
Over a lifetime, the specialisations that many seek will have disappeared. We've moved on from valves and morse code, and we'll move on again from the current mass market methods pretty soon, so some level of generalisation, and importantly, an underpinning level of knowledge and understanding, will be essential for the sucessful engineer.
AGHock, I agree. A talented engineer is a well balanced individual skilled in general knowledge of engineerng fundamentals and armed with special skills in advanced technology. I too mapped out my career having a strong background in basic electrica-electronics knowledge but pursue developing skills in Embedded Systems. Both engineering backgrounds continue to serve me well and have never regret this career path.
I think industry needs both. More importantly each individual should strive to fit those aspects of a career that are in their zone of comfort (augmented by a little sweat to have a working knowledge outside their comfort zone). I firmly believe this to be important if you are to enjoy your working career.
Using any of the profile tests available we fall into many zones. Let's pick two:
The left brained individual (disciplined, task oriented, goal setting, focused, etc) will most likely be happiest in an engineering specialty dealing with as much certainty as possible.
On the other hand the right brained individual (undisciplined, variety thinking, likes many different tasks, somewhat unfocused) may be happier working in undefined areas and dealing with unknowns.
I am from the latter. My lack of discipline was not a friend when struggling with my undergrad Electrical Engineering. Flunked out once, went to another school and graduated in spite of my study habits. Ten years later I completed two M.S. degrees in Applied Math and in Computer Science.
I enjoyed R & D and New Product Development and relished the challenges of uncertainty and solving unforeseen problems. I have had the opportunity to manage many coop students and feel that it is wonderful to see them gravitate to one particular area where they were able to shine when having been rather vanilla in a different area. The coop experience avoids getting that first job in you vanilla area which can be a career true damper.
Being a closet generalist worked out for me. Managing and troubleshooting at companies that introduced among the first electronic typewriters, electronic sewing machines, specialty watches and inkjet printers, including the hotmelt technology that lead to early stage parts manufacturing via 3D printing and "lost wax" parts fabrication. Lots of patents in lots of fields, I really enjoyed my career. Still read to keep up.
For the engineering student, seek out your true interest profile and design your engineering career around it. Get the basics so that you can understand the task needs and read continually outside your specific field to add a generalist capability to your bag of tricks. Then you can go well in many directions!
I think the term specialization is loosely defined in the engineering field. As a ME, I took a controls track in my undergraduate. I was hoping to move into the field of robotics and mechatronics. However, my current skill set has been developed in the field of thermoplastic injection molding (part, process, and tool design). My undergraduate studies in controls have little relevance in my current job (other than occasional automation projects).
Specialization seems to follow the engineers work. I took a job with an injection molding company. From that first engineering job, I have trained and specialized to enhance my productivity in this field of work. As a result of this specialization (nationally recognized certifications), I have a skill set that is highly sought by the injection molding community (I get frequent calls from head hunters). I cannot say this is true for everyone, but just conveying my experience.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This radio show will show what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.