Google’s secret perk? A private hackerspace
hackaday.com
hackaday.com
As for the certification and elitism. The cert process was very straightforward and relevant. People easily forget that metalworking and some woodworking machines don't let you learn by experimenting. These machines kill in seconds, sometimes for mistakes that are very unintuitive. Like wearing gloves has killed several people. Which is unintuitive when you are handling razor sharp pieces of metal. Or the recent fatal accident at Yale where student was pulled in by her hair.
As for the particulars of Ihab. He is a very smart and diligent guy but he had very little practical machine shop experience. He recognized it himself and took classes to compensate. These standards aren't high just due to Rod, but because most Googlers are smart enough to be critical of their own skills when faced with lethal equipment.
Re-posted from this earlier thread on the same subject: http://news.ycombinator.com/item?id=2487979
Edit: s/MIT/Yale/ for the lathe accident. Thanks.
Also, a lot of modern machinists have, even, a basic understanding of materiel science. They might not have taken all the courses to graduate with an engineering degree but they aren't that far off.
This is interesting. How/when are gloves dangerous? I've been TIG welding for about a year now, I self-studied my way through a welding textbook[1] and have friends who are professional welders (union pipe-fitters, etc) and have never heard anything about gloves being dangerous. In fact according to the pipe-fitters, gloves are in the category of eye and head protection of "must be worn on a union jobsite."
[1] http://www.amazon.com/Modern-Welding-William-Bowditch/dp/156...
I think this is a good example of why the safety guidelines are strict and mostly require rote learning. Gloves protect against heat, but everything you have been plasma cutting or welding needs to be handled with tongs because even if the metal doesn't glow it's probably still pretty hot. Most of these rules make sense in retrospect, but they are universally hard to deduce from first principles.
This is very true. I haven't welded since high school, but a single arc weld can get a piece of metal hot enough to burn almost instantly through heavy leather gloves.
I always wear my TIG gloves when operating grinders and was glad I do a while back when my hand made contact with the wheel: what would have been a nasty and painful gash ended up being a scuff mark on my glove.
But that raises another question: Why on earth would your hand be anywhere near the head of an operating milling machine?
EDIT: Thinking about this more, we might have different concepts of "glove." TIG gloves are thin, tight-fitting, and leather.
They also have a "gauntlet type" cut that extends up your forearms and over your sleeves. I don't think they'd be any more likely to catch a machine part than your skin would, and I think they'd actually reduce the chance that a sleeve gets caught.
OTOH, it makes sense that the thick, heavy fabric gloves that are used for MIG or stick would make your hands more clumsy and prone to getting caught in machines. I agree that those are probably pretty dangerous.
Accidents happen. You slip, or get tired and careless and then suddenly your hand isn't where you expect it to be. On a manual mill, it's more likely that your face, not hand will be closer to the bit as you monitor how it's cutting.
A chemistry senior was killed by a lathe at Yale last week working alone late at night when she was pulled against the machine by her hair.
An average size lathe with a 6" chuck will pull 18" of hair in a single revolution. At a moderate speed of 300 rpm that's too fast for a person to react. In the battle of human vs. 3hp motor, human loses. I've seen enough photos of human bodies turned into hamburger by even small industrial machines to not take the danger seriously!
That shop, at the local community college, is the only one I've ever used, so I can't guess what practice is more common.
also when you zero out a mill, you need to touch off a tool (don't know the name) to define your origin spot. it's not sharp and you should spin it at low RPM, but it's still moving, and you shouldn't have gloves anywhere near it.
That is probably the only reason training is so important.
But yes you'd end up hurt either way.
Leather and clothe gloves are not allowed around rotating machinery for that reason. Latex would be okay.
edit: s/Columbia/Yale/
http://www.nytimes.com/2011/04/14/nyregion/yale-student-dies...
This could easily have been any of my peers in graduate school. Lots of us did work late at night, and there are lots and lots of ways to injure or kill yourself with shop equipment. You have to take the work very seriously, be sure you are alert, and whenever possible do your work when other people are around.
Meanwhile: You really, really want the supervisor of your machine shop to be a merciless drill instructor [1] who takes no crap. And you want the people who can't pass boot camp to be sent home without a key to the shop. Why?
A) You, yourself, need more training than you think.
B) Even if you are an expert, you want to be very sure that the person working next to you is trained. You may not be the person who makes the mistake, but you could still be the person who gets electrocuted, or struck by flying parts from a nearby spinning lathe, or crippled when you slip on the spilled oil and drop the hundred-pound vise on your foot [2], or injured by the broken machine that wasn't properly taken offline, labeled, and locked.
C) It's not just the fatal mistakes that will drive you crazy. I've worked in a lot of shared shops. They will waste your time. Waste it by the month and by the year. Because the equipment gets dented, or stripped, or burned out, or rusted, or dirty. It only takes one incompetent or un-conscientious person to ruin something. Entropy is annoying that way.
There's a reason why, e.g., every machinist has their own little box of tool bits and chisels and favorite screwdrivers, and why chefs never let other chefs use their knives.
---
[1] The similarities between good infantrymen or sailors and good machinists (or contractors or modern farmers, for that matter) have been remarked upon before. In all these cases, you're constantly putting your life in the hands of the equipment that's being maintained by your peers.
[2] When your shop instructor gives you the lecture about the steel-toed boots for god's sake get some steel-toed boots and wear them.
If you'd rather discuss this off HN, my e-mail is in my comments.
That said, I have some memory of the shops being fulltime-only. Once you get to Google, go to http://goto/workshops; full details are there.
Edit: If you want to get more info mail me at kai tinkercad.com.
I was at Noisebridge last night and talking about best ways to create STL files for the makerbots there.
Currently, it only supports adding and subtracting geometry - will you be able to pull/push edges/points?
There is no sphere (convex/concave tools), will that be added?
We are thinking about all of those things but some of them might take a bit of time. The internal representation of Tinkercad is a true volumetric format, even simple models would take gigabytes uncompressed, which is why we store them in a cluster. The upside is that our boolean operations are rock solid and the STL:s well defined. The downside is that doing a push pull operation literally means moving the equivalent of digital dirt around.. :-)
Have a dial and text box for saying the reference plane should be "45 degrees to the Z axis"
This would allow one to use the existing tools a bit more precisely.
So, that makes it two of us who are eagerly waiting for the ability to rotate work planes.. :-)
Source: http://www.dailymail.co.uk/home/moslive/article-1367481/Appl... - "Jobs swiftly brought Ive in from the cold, moving the designers into a building on campus and investing in the latest rapid-prototyping equipment. He also beefed up Apple’s security, locking down the design studio to prevent leaks and installing a private kitchen so designers wouldn’t talk shop in public.
I'm personally so much more of a software hacker. But I do have a friend that can do all kinds of hardware hacks, from woodwork to welding. He loves doing it too, but often hits a roadblock with certain ideas since he doesn't have the required machinery and is currently employed outside this field of work. And without the funds, also doesn't have the opportunity to go very far with his hobby.
Unfortunately, he also doesn't have the software background to work at Google. It's an amazing perk they offer for sure. It'd be really cool if they expanded this to allow some non-computer geeks to pitch in every now and then though.
"which is open to any employee that meets some pretty strict requirements. A written test is given before an employee can access the facilities, and even then they must be deemed worthy of working on particular pieces of equipment"
I'd like to assume this is a safety issue, but this is Google which is known to be pretty elitist. Some employees will always get more perks than others (more money, more vacation time, more options...). Part of the very nature of picking "haves" is that you inherently devalue the "have-nots".
One step forward for having a more streamlined interview process, one step back for having interviews/tests for access to perks even after you're employed.
I be more concerned if google let any employee walk in and pick up a torch. If you want to be paranoid not having certification requirements would be evidence of some scary social darwinistic elitism.
I work at a company that produces lithography machines, and to get access to the clean-rooms you have to do all kinds of extra courses and tests, depending on the kind of technology used in there.