Kevin, you make a good point. One needs to understand the underlying science and mathematics of a systems to be able to use it effectively in situations where the software might have a problem (or when the senesors go out). It is also useful when transitioning from one system to another.
Manny years ago, when I was the CIO of a spacecraft plant, we decided we needed to purchase our own system for running engineering codes. Until then we had been using the mainframe that belonged to finance. That machine was great, by the way, and they had purchased a vector processing unit that greatly speeded up our engineering codes. The issue was purely financial and control. During this whole process (and finance made it a torturous process), the head of Engineering (a great guy with a PhD) asked, are we becoming too reliant on computers? One of the older managers from the mechanical engineering group (our biggest computer user) answered him. Before we started using FEM codes heavily we had 200 engineers in the stress group. We were working on two projects at a time. Now we have firve or six projects going with 25 engineers. Question answered.
I would hope that the engineers would know the physics and mathematics underlying the tools they use. The situation you mention is interesting becuase I could turn the argument on its head. I think it is likely that the custom software an engineer might write to compute a value would be more likely to have errors. The package software, while not perfect, has many users, who stress it heavily. It also has a number of people tasked with fixing errors and a quality control process that looks for problems. That is what you are paying for when you pay support fees.
Maybe I am seriously behind the times here, but there are automated systems for parallel parking?? I have always done it myself...so I suppose I could easily take over if one of these systems failed. That said, my car is too old to have such a system...but I'd be fascinated to know more about them!
I am amazed that many readers seem to discover this kind of system. It has been around in Europe for at least ten years and is very common, especially in compact and subcompact city cars. I own a Mercedes Class A of 2009 with this equipment and it works very well : it detects an empty space (right or left in one-way streets) while driving at less than 20 km/h and manages the steering for the parking; You just have to switch between D and R and brake according to the instructions. The system disconnects as soon as you touch the steering wheel (It is usually installed on cars with electrical power steering).
Accuracy is quite good and you end up parked along the curb with an offset of 6 inches +or- 1 inch...
"In engineering, there is much more at stake than simply losing a parking place."
I know parking system is more on technicalites but I believe we can resolve parking problems manually. In America, parking tickets are too insignificant for out budget-strapped police departments to devote much effort to enforcing. For many drivers, they have become little more than an annoyance, entitling them to park in any manner they please just to save walking a few steps. But apparently that is not the case the world over, as two recent cases in the U.K. illustrate.
Interesting, I got my first drivers license while living in Germany and back then if you couldn't parallel part you didn't get your license (as well as a few other parking manouvers). At that time I didn't know the math involved so I don't quite understand why this is used as a introduction to why we as engineers need to know the math behind technology systems. I still park within a few inches of where I want to without considering the math and will continue to do so. When I got my first car with ultrasonic park assist sensors (not actual parking, just beep before contact) I used it for about a week before I got so annoyed by this system that would tell me I was about to hit when I knew I had at least an inch or 2 to go, so I turned it off never to be used again. Really to the point (for parking) it's a matter of hand eye coordination, and while the math is certainly relevant to a computer performing what is a basic task for humans, it plays no part in how people solve the problem. I would wager that it's possible to have an engineer that can do the parking manouver on a calculator effectively but would fail at the actual task due to the lack of hand/eye coordination for the same basic reasons that an engineer could write a book on how to solder but may do rather badly at it (as most engineers I know seem to) :-) Machines just simply solve problems differently to the way we do, we only need the math to tell a machine how to do it.
Naperlou: I believe you've identified a point which I've been railing on about for quite some time. IF a technologist is not steeped in the fundamentals, then to me it's like building a skyscraper on sand...... it won't last long. It seems that many of our institutions of higher learning have eschewed these fundamental concepts in order to present the "latest & greatest", much to the detriment of the ultimate goal. That's why I think colleges & universities would be better served in offering a course in Slide Rule operation, as opposed to a recent blog about Soldering Techniques.
What if you don't have or have never had a car with an automated parallel parking system?
The metaphor is somewhat broken as it's not really a capability everyone and their cars possess, but it can go along with the metaphor. You do without, and do it yourself. In an attempt to show how technology can fail, you assume our complete dependency on it. If the tools exist and are affordable you use it. But it's always important to know what's going on at the core. The layman might not know what a specialized tool is for, but the specialist using the tool has usually been taught what exactly it's doing. Don't most US states require a parallel parking procedure as component of a complete driving test?
Dayling, recently I had seen a YouTube video about automated parking. Driver can leave the vehicle at parking ground and the car will found out free parking space with the help of satellite. Finally the self driven mechanism will drive and park the vehicle at designated slot. I think such proto type is developed by GM
I believe that VW has run TV ads showing just such a system. And, it seems to me that I've read where there are several independent companies which provide this technology. I also believe I've seen a video from a German company which describes this action, and there's a company in FLA which also designs & builds these systems..... ROBOTIC PARKING SYSTEMS, or similar name. I don't recall exactly.
:-) It said even the worst driver..... I took a picture last week of a driver that ran off a cliff backing down his/her driveway, so maybe not :-) but certainly innovative. There's also a 4WD electric car that has 4 wheel steering for similar effect. Amazing the effort people go to to park a car.
The fact is that I never have had a car with an automatic parking system, and I would not want any of them that I have seen in the past 20 years, even if they were free. Part of my driving test when I got my license was to parallel park, and I passed it well. That was 50 years ago.
But if somebody lacks the insight and understanding to do some task then they are completely at the mercy of others to do it for them, or go without. And I see drivers almost daily who should not even be driving down a straight road, mostly because they lack the ability to pay attention to what they are doing, at least, not pay attention enough to drive safely. And I wonder how the automatic parking system handles exceptions, such as a pothole next to the curb. Does anybody know ?
The same concern is valid about a lot of engineering tools. As Bob Pease used to remark frequently, the simulation tools only deliver a quality that is at best as good as the model used. But if the model is not correct then the results will probably not be right either. They may be "good enough", but in that case then regular thinking should be able to provide as good an answer.
Relative to all that, way back at school we were told to apply the "Maselowski criteria" to analysis results. That criteria was the question, "Is this answer reasonable". That still holds true today. He may not have made a big point about that, but it has served me very well, and led to checking results prior to announcing them to others. Private math errors are much less embarassing.
Kevin, there is no doubt that parking is a big issue with most of the cities, especially with metropolitan cities. Multi-level parking is the only solution and in future it won't be sufficient to accommodate the traffic. In such a scenario, we won't be able to image even about parallel parking
It might be more important to consider if the "automated system" that parks your car had done so properly. Computer aided design tools are only useful if the user can recognise when they are being LIED to. Many times I'll run a Spice simulation and ponder why it did not give the expected result. Did I do something wrong, or did the program fail to get the answer I expected. Even if a user does not have the technical ability to do the task manually (hey, solder is my backup plan) they have to realize where the pitfalls lie. Too often this does not happen as users don't understand even basic principles, let alone rigorous analysis.
How about... What if your calculator couldn't do "long division"... or even "square roots"
Great insights, HarryB. Mechanical design CAD system's Sketcher algorithms are often infuriating as they make inaccurate relation assumptions. I wrestled with them daily. Your point is clear: You have to already know what's right, and let the computer assist you in getting there a little quicker; Not just do it for you. ( I wonder if I can still do long-hand division-?)
Great post, Kevin. Just as young drivers should learn to parallel park their cars manually, engineers should be required to know the the physics of their particular products before they start making extensive use of computer tools. Sometimes, it can take years before an engineer fully grasps that physics. An engineering degree serves as a great foundation for learning, but nothing replaces experience. Good companies (and good senior engineers) will help new engineers gain that experience. It can sometimes be tempting for companies to circumvent that process, though, given the cost and speed advantages that computer tools provide.
I agree, Nadine. Parallel parking is a bit of a lost art these days. None of my kids are able to do it. I think it's interesting that Kevin chose to make his point with a comparison to parallel parking. There is indeed a "parallel" there.
I totally agree. I love to create designs by hand because it allows me to really engage with components. As I draw circuit schematic designs, I have this internal talk on the electrical behavior between a resistor -capacitor network and does the combination makes since for the intended design. Yes, engineers have become too dependent on digital tools and the mere though of going back to using manual instruments does scare them.
I agree that the technology might benefit people with weak eyesight but that is not a fix in itself, it could produce serious results for people not fully fit to drive their vehicles on the road. The government and concerend authorities and organizaitons might think this is their sinere effort, but I am not in favor of it as an individual arsenal thanks
Good point about the experience. I stopped hiring CS majors (no offense folks) for writing firmware because they rarely had experience with the limited resources in an embedded system. Huge structures would be passed back and forth on the stack and so many Interrupt Service Routines would be running that the processor never had a chance to complete a task before the next IRQ. The software was well-written and easy to understand, it's just that the author's lack of experience on an embedded system would gobble up every machine cycle and byte of memory.
I guess that I am in the wrong socio-economic class or something, since my 2011 Subaru Forester doesn't have anything like automatic parking or that nice toy, a rear-vision video display or even a sonar rangefinder to tell me where the other guy is. Are there really people out there who don't know how to parallel park (at least in theory, even if they try to avoid it)? When I got my driver's license (in Massachusetts, 1966), parallel parking was part of the driving test; you definitely had to be able to at least muddle through...
At this year's MD&M West show, lots of material suppliers are talking about new formulations for wearables and things that stick to the skin, whether it's adhesives, wound dressings, skin patches and other drug delivery devices, or medical electronics.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.