4 ms·

I came here to say something along these lines. Letting accountants drive engineering is a fatal mistake, but so is engineering without any thought to costs and

I came here to say something along these lines. Letting accountants drive engineering is a fatal mistake, but so is engineering without any thought to costs and profits. I’ve walked away from more than one startup whose business plan was, “Let’s design the thing, and only after that will we think about COGS and competitive pricing.” That way lies madness, and Juicero.

geophile 7y ago
Juicero adds another function to the discussion: marketing. Juicero was an example -- a very comical example -- of marketing over everything else: technology and bean counting. Yes, a balance among all three functions is necessary. However, the 737 Max case is definitely about letting bean counters dictate engineering choices: saving design costs by pulling the ancient 737 design off the shelf; the software fixes to the problem introduced by those engines on that frame; and then pretending that the plane was just like an old-school 737, hiding all the complexity and compromises from pilots to avoid recertification costs. And even the decision to charge extra for redundancy of the critical sensor driving the software.
- PhasmaFelis 7y ago
  Juicero may also have had the opposite problem: the dangers of letting engineers play unsupervised. It appeared to be, and maybe should have been, a cheapjack gizmo swaddled in marketing glitz to justify its ridiculous price. But someone did a teardown and discovered that it actually did contain a very expensive, precision-engineered, completely unnecessary two-ton press. To squeeze fruit pulp out of a tube.
  - mikepurvis 7y ago
    The legendary Juicero teardown: https://blog.bolt.io/juicero/ https://blog.bolt.io/juicero/
    - OnlineGladiator 7y ago
      This one is a lot more fun: https://www.youtube.com/watch?v=_Cp-BGQfpHQ https://www.youtube.com/watch?v=_Cp-BGQfpHQ
      - quakeguy 7y ago
        That one was a field day. Hate him, or love AvE, this vid is just fun to watch again. Thx.
      - Tsiklon 7y ago
        AvE's style is very entertaining. I always end up coming away with understanding a little bit more about physical engineering and manufacturing after watching something of his.
  - dsfyu404ed 7y ago
    And all the bits around said overkill press were designed from flat stock with features cut in it to hold things. It was reminiscent of all the terrible "we only know a few manufacturing techniques and the world if full of nails that we will hit with those hammers" type designs that engineering undergrads who have no real world experience crank out. I say that as someone who used to work for the machine shop for a ME program. They probably could have cut the cost by a huge amount by using an aluminum casting (though with more engineering effort they probably could have used plastic but that would have upped the tooling complexity and I dunno if their planned volume would have made that worthwhile) and machining all the features to hold the moving bits into it (bonus points for doing it all with one tool and one holding fixture) and designed their press mechanism around the gearbox of some power tool you can find at Harbor Freight for under $50. New engineers being new engineers I would put serious money on all their machining having stupidly tight tolerances. New engineers love doing that for some reason.
    - celim307 7y ago
      > New engineers being new engineers I would put serious money on all their machining having stupidly tight tolerances. New engineers love doing that for some reason. Im a mere software engineer but I see a lot of over engineering in new grads in the field as well. I sense it comes from either wanting to dick wave, or fear of being thought of as too amateur.
      - PhasmaFelis 7y ago
        I think it comes from wanting to do things "right." Sloppy design is highly visible and frustrating to an engineer (hardware or software), and we want to avoid that. It takes some more experience to realize that hyperfocusing on a single optimization misses the big picture and hurts other important factors.
- rainbowzootsuit 7y ago
  The "AOA-Disagree" display, as far as I understand, isn't an enhancement to the redundancy of the AOA sensors. That would be different software for the MCAS. It is just a display of the AOA disagree condition, and MCAS operation is the same.

geophile 7y ago

Juicero adds another function to the discussion: marketing. Juicero was an example -- a very comical example -- of marketing over everything else: technology and bean counting. Yes, a balance among all three functions is necessary. However, the 737 Max case is definitely about letting bean counters dictate engineering choices: saving design costs by pulling the ancient 737 design off the shelf; the software fixes to the problem introduced by those engines on that frame; and then pretending that the plane was just like an old-school 737, hiding all the complexity and compromises from pilots to avoid recertification costs. And even the decision to charge extra for redundancy of the critical sensor driving the software.

PhasmaFelis 7y ago
Juicero may also have had the opposite problem: the dangers of letting engineers play unsupervised. It appeared to be, and maybe should have been, a cheapjack gizmo swaddled in marketing glitz to justify its ridiculous price. But someone did a teardown and discovered that it actually did contain a very expensive, precision-engineered, completely unnecessary two-ton press. To squeeze fruit pulp out of a tube.
- mikepurvis 7y ago
  The legendary Juicero teardown: https://blog.bolt.io/juicero/ https://blog.bolt.io/juicero/
  - OnlineGladiator 7y ago
    This one is a lot more fun: https://www.youtube.com/watch?v=_Cp-BGQfpHQ https://www.youtube.com/watch?v=_Cp-BGQfpHQ
    - quakeguy 7y ago
      That one was a field day. Hate him, or love AvE, this vid is just fun to watch again. Thx.
    - Tsiklon 7y ago
      AvE's style is very entertaining. I always end up coming away with understanding a little bit more about physical engineering and manufacturing after watching something of his.
- dsfyu404ed 7y ago
  And all the bits around said overkill press were designed from flat stock with features cut in it to hold things. It was reminiscent of all the terrible "we only know a few manufacturing techniques and the world if full of nails that we will hit with those hammers" type designs that engineering undergrads who have no real world experience crank out. I say that as someone who used to work for the machine shop for a ME program. They probably could have cut the cost by a huge amount by using an aluminum casting (though with more engineering effort they probably could have used plastic but that would have upped the tooling complexity and I dunno if their planned volume would have made that worthwhile) and machining all the features to hold the moving bits into it (bonus points for doing it all with one tool and one holding fixture) and designed their press mechanism around the gearbox of some power tool you can find at Harbor Freight for under $50. New engineers being new engineers I would put serious money on all their machining having stupidly tight tolerances. New engineers love doing that for some reason.
  - celim307 7y ago
    > New engineers being new engineers I would put serious money on all their machining having stupidly tight tolerances. New engineers love doing that for some reason. Im a mere software engineer but I see a lot of over engineering in new grads in the field as well. I sense it comes from either wanting to dick wave, or fear of being thought of as too amateur.
    - PhasmaFelis 7y ago
      I think it comes from wanting to do things "right." Sloppy design is highly visible and frustrating to an engineer (hardware or software), and we want to avoid that. It takes some more experience to realize that hyperfocusing on a single optimization misses the big picture and hurts other important factors.
rainbowzootsuit 7y ago
The "AOA-Disagree" display, as far as I understand, isn't an enhancement to the redundancy of the AOA sensors. That would be different software for the MCAS. It is just a display of the AOA disagree condition, and MCAS operation is the same.