Loved the premise, the execution left me a little cold. There was much here that was interesting but then it would start turning into a text book. I'm sure that for an engineering student or a doorknob designer an analysis of the eight or nine distinct actions and decisions involved in opening a door are probably of use. For me, less so.
I still think its worth a read. I liked the really excellent argument that much of what is termed human error is in fact poor design. And that one sign of a well designed product is that you don't need a multi page instructional manual to figure out how to use it. There were lots of interesting things like those in here, I liked the idea that designers should plan for operator mistakes and wherever possible make them easily reversible. I liked the point that almost no truly new product or idea really works all that well on the first try because hey, its new nobody knows how yet! So if we want innovation, we have to give new ideas some room to fail a time or two.
There is a whole lot in this book that is really interesting, but it does suffer from a confusion of tone. Is it a textbook for engineers and designers or is it for a general audience? It teeters back and forth and never does really decide. But I guess if I want innovation, I have to give this book some room to be dull in spots.