Simplicity as an advantage of a certain method or approach to a problem

[Axe199]

hi, i don't know if this qualifies as a question or not, but this is something i can't really understand.

i am an engineering undergrad, when we are learning about the pros and cons of different method of doing anything , for example maximum power point tracking in PV systems, we discussed a method called perturb and observe, and the main and most important advantage of this method was simplicity but has some serious drawbacks.

The part i don't get , why is it still used, i mean, don't the guys designing these things have years of experience and sometimes masters degrees? surely they can understand the other more complicated better methods. There are other examples for this but this is the one i can remember right now :D.

So to sum up my question, why is simplicity considered an important advantage in industry when it's easily overcome with proper training ( or at least i think so)?

 

[scottdave]

You will often find in engineering, that sometimes the more complex "better" method may not be all that much better. For example, what costs are involved in using the more complex method: processing speed, material cost, other resources? How much gain is associated with these additional costs?. Are the gains worth the additional costs?

 

[jrmichler]

This is a very good question, and one that more engineers should ask.

Because the real world is very much different from the mathematical world. In the real world, nothing works perfectly all the time. An MPPT controller that consistently delivers 97% of the theoretical peak power is better that a controller that delivers 99%, but only works 90% of the time. Simple solutions are easier and faster to design, easier and faster to develop, easier and faster to get into production, and easier and faster to troubleshoot and repair. Also lower cost, therefore easier to sell.

In the end, it gets down to a tradeoff between cost and performance.

 

[jim hardy]

Complexity adds failure points.

http://rigo.altervista.org/papers/engineering_procedure.pdf

 

[berkeman]

Good question @Axe199

To add to the good responses so far, there are usually many different things to consider when optimizing a solution in a product design. Reliability, size, maintainability, cost(s), etc. will need to be considered.

Also, in circuit design, you need to think ahead to what the life of the product line will be, and consider possible obsolescence of the parts you are specifying. For example, you may have a choice between a slightly cheaper design that uses a nice combo microcontroller with some widget you need (DSP, ADC, whatever), versus using separate components for that part of the design. If the nice combo IC is made by a smaller company that does not have a long track record yet, and the alternative parts are pretty jellybean in the industry, you may decide to go the safe route with the slightly more expensive design that is easier to support over the long life of the product.

Where I work now, we have several legacy products that still sell very well, 15+ years after they were designed. We have to spend way too much time finding replacement parts for ICs that are going obsolete now...

 

[jim hardy]

Where I work now, we have several legacy products that still sell very well, 15+ years after they were designed. We have to spend way too much time finding replacement parts for ICs that are going obsolete now...

We actually ran across some memory IC's with tiny ten year batteries inside them. They were in a fuel handling crane for the nuke plant.
I guess some young designer didn't realize we nukes expect stuff to last forty years.

 

[Axe199]

hi, THANKS VERY MUCH for all the very useful replies.
I talked to my professor about this and he made two good points:
" the fact that you have an engineering degree doesn't mean you can't make mistakes, a complex design will often have design mistakes and some time these mistakes won't be discovered for a long time by then it could cause a major failure" i guess that's the same point @jim hardy was making.
2- " the complex design have problems but often when this is a new product supposed to compete with other products from other companies they will tend to hide any problems they couldn't find a solution too :) "
sorry for the late reply

 

[scottdave]

...they will tend to hide any problems they couldn't find a solution too :) "
sorry for the late reply

Nobody does that, do they? Ha.
Reminds me of my Engineering Ethics class.

 

[jim hardy]

Nobody does that, do they? Ha.

In IT industry ? Naaaahh...

 

[Klystron]

[snip]
So to sum up my question, why is simplicity considered an important advantage in industry when it's easily overcome with proper training ( or at least i think so)?

Along with ethics, simplicity provides a guiding principle for your decisions as an engineer.

Often after learning a new methodology or training in new technology we want to incorporate the new knowledge in designs. At the same time we each have favorite methods to use to solve problems. Simplicity offers guidance. Choose the least complex methods that best solve the problem and fulfill requirements within budget.

Most projects require team effort. Simplicity provides a common structure, a unifying goal for team members. Spark ingenuity, invite creativity, but Keep It Simple (KIS).

 

[anorlunda]

I can't believe nobody mentioned KISS. It is the most cherished of all engineering principles IMO.

https://en.wikipedia.org/wiki/KISS_principle#In_software_development

"The KISS principle states that most systems work best if they are kept simple rather than made complicated; therefore, simplicity should be a key goal in design, and unnecessary complexity should be avoided. " Note especially the word unnecessary.

See if you can get this book from the university library.
Computer-Related Risks
by Peter G. Neumann | Oct 18, 1994

It is a very enjoyable book. The examples in the book make it clear that KISS violations are the root cause of very many real world risks.

 

[Rive]

So to sum up my question, why is simplicity considered an important advantage in industry when it's easily overcome with proper training ( or at least i think so)?

Bit of a late response for a thread more than year old, but let me provide an example.
Once upon a time we had a very simple release strategy: 'pack up everything as required by the manufacturer'.
Since the manufacturer tended to complain if something went wrong, it worked perfectly. Even newbies got it right fast enough.
Later on, the BOSS had a brilliant idea: let's make an automatic release script so it'll be uniform, fast and easy!
The result is, that right now the BOSS is the only one who understands all the magic involved with the parameters, sequence and name policy required to adjust to all the differences/exceptions required by the different manufacturers, so the release is now done by him alone and if he is on leave, we're stuck.
Simple, isn't it?

 

[Klystron]

Bit of a late response for a thread more than year old, but let me provide an example.
Simple, isn't it?

Odd, this thread popped up as new when I first logged into PF5 Saturday. Perhaps because it was one year old to the day? Good answers in any case .