« What do you want to do with your life? | Main | Pendulum »

To Engineer Is Human

I'm thinking that while I'm posting comments about these books, I'm not really going into depth. It's seems to act more as an outline of what I'd want to talk about with someone after I'd read the book. Since most of the books that I've been writing up came from the Museum of Science Book Club, that discussion has been taking place, just not in an electronic format. I think I tried to summarize one of those discussions, but felt it was just reduced to sounds bites.

If nothing else, I find that by putting these comments together I have a better chance of actually remembering what the book was about and hopefully recalling some of the details. It is amazing that for all that I read, I usually don't assimilate that much of it consciously, it instead filters in and sometime in the future it begins to make sense if I come back to it. I'm particularly finding that to be the case with my current book. I read an entire chapter today over lunch and knew that I would have to read it again to understand what it was trying to get at. Now onto this book:

�To Engineer Is Human� by Henry Petroski focuses on the need for failure in engineering to keep innovation in check and provide a basis for new ways of thinking about safety and maintenance. Innovation requires risk, it is the job of the engineer to think about those risks and ask the right questions [201,222]. If all engineers played it safe the rate of innovation would slow to a snails pace or the cost would remain too high. The risks are controllable if they are within reasonable bounds of existing experience [5,201]. The book covers many examples of well known failures, the failures being well known since they are the anomaly [106]. The failures discussed include the walkway collapse in the Kansas City Hyatt Regency Hotel, the Tacoma Narrows Bridge, and the AA DC-10 crash.


Some of the main points from the book:

vii: Engineering is about design which is creating something that hasn't existed before.
viii: Failures in engineering lead to greater safety margins and new periods of success.
5: The risks of technology are controllable. After the DC-10 crash 10 million flights carrying over half a billion people flew without incident.
19: The three little pigs in an analogy of structural failure in the face of adverse conditions.
25: The original Speak & Spell had a defective keyboard that broke after awhile. However, since all the users loved the machine many worked around the problem.
27: More thought of safety is given to devices that have a higher consequence of failure.
34: It is important that we learn from the mistakes of others.
45: Engineering success is one of high probability, not certainty.
48: For new designs to succeed the ideas need to be extrapolated instead of copied since what makes a design work for one situation may not apply to another.
56: Success in the past was more trial and error instead of actively thinking about what could and couldn't fail. In general the cathedrals where most likely build using pre-rational structural engineering.
61: The definition of an engineer used to be related to designing, but was muddled when those same people also started driving railroad engines.
62: Many structural failures, are failures of the mind to consider what could have caused it to fail.
62: While it would be great to remove failure, it would be impossible to go beyond the state of the art without it.