Join 3,380 readers in helping fund MetaFilter (Hide)

Engineering toolbox and the design process
February 25, 2010 8:03 PM   Subscribe

What are some good places to find concrete examples of people using computers / math / engineering tools to solve (difficult or easy) problems in an intelligent manner?

Coming from a engineering background, you are taught a varied design skill set which can be used. Most of this is in theory.

I'm looking for industry examples which are shown or described in way that shows these tools/software being used. This could be as big as the full design process used, or a specific point in a larger design. Where can I find online resources for this sort of information.

I'm looking for media online / presentations / papers / websites / etc..

Another side question, what skills do you use for your job?
posted by MechEng to Work & Money (11 answers total) 3 users marked this as a favorite
There was a recent FPP about using R for data analysis. You might also not have heard of the classic The Mythical Man-Month or The Soul of a New Machine.
posted by axiom at 8:09 PM on February 25, 2010

Henry Petroski is a sort of "you love him or you hate him" guy but he writes an awful lot about engineering in a somewhat-readable way. I enjoyed his book Pushing the limits: new adventures in engineering which talks about very specific design/build projects all of which took engineering ideas in a new way [i.e. they did something that people hadn't done before and so were heavy on theory and a lot of "hope it works" idealism]. The book talks a lot about how people made the calculations about whether the projects would work or not. And sometimes they did and sometimes they didn't.

Side note: Soul of a New Machine is about my dad's project at Data General, so this may be a topic near and dear to me. His background is in astrophysics which is another area I think you see people doing a whole lot of math to arrive at one [hopefully] elegant result.
posted by jessamyn at 8:16 PM on February 25, 2010

I'm not entirely sure it's quite what you're looking for, but you might want to have a look at Computing in Science and Engineering / Computing Now from the IEEE. It has a information-processing focus.
posted by oostevo at 8:24 PM on February 25, 2010

The Bombe
posted by zombieApoc at 8:28 PM on February 25, 2010

Monday through Friday I develop Host Cell Protein analysis in the wonderful world of bio-pharmaceuticals. Because it's not a single analyte it tends to be an ill mannered assay with lots of quirks. A few years ago I started trying to model these mathematically since the traditional method (vague hand waving) was faltering under the data.

Since then the skill I've used most is oratory since it's been an uphill battle to convince my biologists coleagues that proteins follow the same laws of physics as everything else.
posted by Kid Charlemagne at 8:29 PM on February 25, 2010

Real life heuristics for the Bin packing problem and other NP-hard issues come to mind. I remember reading a paper a while back about a tweak that was made to FedEx's bin packing algorithm that saved them millions of dollars but can't find it now -- hopefully someone can fill in the blank.
posted by telegraph at 8:42 PM on February 25, 2010

Have a look at the technology-related videos from the TED conferences. You'll find dozens of examples.
posted by qxntpqbbbqxl at 10:28 PM on February 25, 2010 [1 favorite]

What are some good places to find concrete examples of people using computers / math / engineering tools to solve (difficult or easy) problems in an intelligent manner?

If you consider use of CAD to fulfil those requirements, almost all engineering projects of any complexity make a lot of use of CAD. CAD companies' websites are loaded with customer success stories.

Or, if you're interested in antenna and lens design, this is an interesting article.
posted by Mike1024 at 12:23 AM on February 26, 2010

I use a DSO for diagnosing electrical and mechanical malfunctions in automobiles. Use it for checking sensor signal integrity and synchronization, I have a number of inductive pickups that help me measure current flow through circuits and analyze the function of electric motors and pumps through waveform produced by their current draw. I also have inductive pickups for measuring the magnetic field and high voltage spikes output by ignition systems. I can use those waveforms to determine the integrity of the ignition system as well as some combustion characteristics which can change the resistance of the circuit. Since the spark gap is in the combustion chamber, I can look at the characteristics of that spark or burn and determine a lean condition or low compression which will change the demand on the ignition system.
I can also use a pressure transducer for my scope to help determine mechanical malfunctions in systems that I'm not inclined to disassemble. With a multichannel scope, I can directly compare fuel injector actuation with fuel pressure fluctuation to determine injector integrity and consistency without removing any components. I can also measure cylinder compression and leakage with a pressure transducer.

So, my labscope is an awesome tool for diagnosing cars. It's like an X-Ray machine that can help generate a picture of what's happening in the car without taking anything apart.
Over the past couple years, effective labscope or DSO use has become increasingly popular in the automotive field among some of the more savvy shops.
posted by Jon-o at 4:34 AM on February 26, 2010

Jon-o, that's fascinating. Are there any books on the topic? I've only ever gotten as far as tapping into the ODBIII connector.

As to the original question, one of the better sources of such information with respect to software is their marketing literature. Look for videos of the AutoDesk siggraph presentations or other such demos -- they show specific instances where their product integrates into some pipeline and solves specific problems.

One story that comes to mind is the Quake Fast Inverse Square Root.
posted by fake at 8:44 AM on February 26, 2010

I have had a chance to read one of Henry Petroski's books, I can't remember the name but it was about the evolution of designs of simple objects over time, such as the fork or knife.

TED talks is definitely a good source of new technology, although sometimes not from a technological standpoint. Hans Rosling takes the cake for best presentation of a data set found here.

The CAD success stories are very interesting as well. If you have any more of those I'd be interested in seeing them as well. Other important features.

A friend directed me to the site Tucows a while ago. This site is useful in finding freeware software. A search here led me to Enguage Digitizer which was very useful in turning a scanned graphs, or picture graphs into usable data. I then took the data points from the software plugged these into excel via a csv file. I the used linear interpolation on the graph to use it in needed equations.

Another useful piece of software I've come across is Karen's Directory printer. This prints the list of file names, folders, information, etc. from your computer drive.
posted by MechEng at 9:22 AM on February 26, 2010

« Older what to bring to dinner for th...   |  The layman's guide to data bac... Newer »
This thread is closed to new comments.