This one hits home for me... My apologies for being harsh, but this is a real world we live in, not a dream one. The one thing I must mention however, is my opinions are based upon my experience as a startup, and as a small manufacturer. In a large organization, I would assume the expectations would be different.
I've hired a fair amount of people, including seasoned folks and recent grads over the years. I think there is one thing missing from the discussion. As a business, we are in business to make money, we are not educational institutions, we are not charities, we have stockholders to keep happy, we do not have the time to be altruistic.
We do expect the newbie to hit the ground running. If we did not, we would hire liberal arts grads, or even folks without educational credentials and train them. I hired a violinist/electronics tech years ago and trained him. From a creative standpoint, as well as a technical standpoint, he was a lot more capable than most senior level engineers in short order. Still, the training overhead is high, but in reality, not much higher than some entry level engineers with a BSEE. The key imho, is not vast technical competance, but the ability to learn, and to learn quickly.
Now as far as expecations go for newbies. Its pretty easy to determine whether a candidate is on the ball and capable or not during the interview. If I am hiring someone, the basics better be rock solid. More than a few times, I gave them some simple circuit issues to solve, only to come back, and find they were clueless. Thats pretty disheartening. I'm not going to teach basic circut theory to a 4.0 BSEE who states, I just memorized that for the test, I never knew I might need to use it.
By the same token, I don't expect the newbie to know my CAE tools, business processes, regulatory issues, or manufacturing technology I do expect them to take the books home and study to get up to speed. I also expect a newbie to put in long hours, much of it devoted to learning these peripheral issues. That is where a good program with substantial non-engeering course work enters the picture. The ability to learn quickly is often times a greater asset than being a guru at theoretical engineering calculations, at least from my perspective.
The other issue I have is the overly simplified and overly complex methods that are used in education. Many students become masters of the complexity of solving differential equations from a theoretical math standpoint. Real world problems have real world effects that do not fit into easily solvable mathematical processes. If I take a real world circuit, and try to theoretically analyze it, I will remove variable, after variable until it fits within a nice solvable math model. SPICE macromodels are probably the most common example of fitting the part to the math instead of the other way around which is why I rarely use them. If I add in real world effects, the math model becomes so complex, it may well take a poor student a whole semester to analyze it, and even then, more than likely he will get it wrong.
By the same token, there are many scenarios where a slide rule is a huge time saver. We don't need to dwell on the minutia, often times a ball park is more than adequate.
I'm not saying we need to throw theoretical math out the window. What I am saying is that its a small part of the solution, and that real world problem solving is drastically different. I'd rather see one less semester on calculations, and one more semester in the lab, where things physicially happen. My ideal analog lab course, if I were to make one would include how to use a simple high gain op amp circuit as a force sensor, humidity sensor, temperature sensor, humidity sensor and an airflow sensor, and then an attempt to model such a part with these real world phenomena. Note, I am not saying to interface an op amp to a sensor, but to use the actual op amp as the sensor itself. I'd also include a section on Smith charts, nomographs, and slide rules such that students could see the ball park methods as well, despite the fact that many of those methods have been superseeded by a computers. Lastly, the lab course would involve troubleshooting and repair. Its a rare prototype that works perfectly on power up. The exposure to real world noise mitigation, tracking down glitches, solder shorts, and open traces, combined with situations where one is sure that Mr Maxwell and his equations were wrong would add a lot of value to the experience. A newbie doesn't need mastery, but he should have at least a minimal exposure to the real world. A single team oriented senior design project usually falls way short of such exposure.
As far as co-ops and internships go, its pretty difficult. Usually the intern gets grunt work a fair amount of the time. Occasionally they get some real experience, and sometimes the fortunate are able to get in the trenches. Its better than nothing for sure, but too many times, its way too easy to put them in a corner doing grunt work. I know, I'm guilty of it, despite trying very hard to include them in the day to day activities. When a customer has a problem, that's the priority, not the intern.
After my 2nd intern, I did come up with a solution. When I got too busy, I'd put him to work as a tech. I'd have him build cables, solder prototypes, work with the machine shop guys, and troubleshoot various low priority circuits. Granted, most engineers do very little tech stuff... but it is a skill we need to pull out of the closet from time to time, and it does take a while to learn. After the internships, I'd often get comments as to the psotive experience of being exposed to the technician work. Its a whole lot easier learning that side of being an engineer as an intern, than having to learn ones soldering skills in a motel room the day before a big tradeshow.
Ron
For example, in the problem you all seem to be talking about, it sounds like the numbers used for the example would lead to catastrophe, so have some fun and after giving the exercise for practice and to demonstrate the theory of how to use the equations and principles, toss in a question for a bonus point: What would happen if you really designed a system this way? Get the students to think about reality a bit.
