• (disco) in reply to anotherusername
    anotherusername:
    You may not have that space; your robot still has to interact with the rest of the real world, which can't be scaled up in size.

    still waiting for a concrete example of a situation where this is the case. you have not provided.

  • (disco) in reply to accalia

    You can't imagine a situation where "hella complicated", "complexity would rise in a superexponential manner" might put you outside of the allowable design requirements, even given unlimited time and money? Which, by the way, I thought was plenty generous, given that you're intentionally ignoring what's physically possible for all practical purposes?

    Fine. You already agreed that, for practical purposes, it's totally impossible. So this argument is really a moot point.

    At this point I feel like the burden of proof is on you to show that, for every imaginable system, a physical interlock would be possible for every motion such that it's failsafe. That's a rather incredible claim that you've made.

  • (disco) in reply to anotherusername
    anotherusername:
    Fine. You already agreed that, for practical purposes, it's totally impossible.

    no. i said for practical purposes it can be impractical, even extremely impractical.

    it is not however, impossible.

    you might not be willing to spend the money or time to implement it, but that does not make it impossible. Design requirements restrict the problem space and in any case are most useful as a description of what the customer wants, they often don't care how you get from A to B so long as you do.

    I stand by my statement. A system of physical interlocks is possible and in cases where we're dealing with the safety of humans that is a very good thing indeed.

  • (disco) in reply to accalia
    accalia:
    impractical
    accalia:
    not however, impossible
    [image]

    "Impractical" and "for all practical purposes, impossible" mean basically the same exact thing. Especially when calling it "impractical" is more of an understatement than it is even the tiniest bit hyperbolic.

  • (disco) in reply to anotherusername
    anotherusername:
    accalia:
    impractical
    accalia:
    not however, impossible
    [image]

    "Impractical" and "for all practical purposes, impossible" mean basically the same exact thing.

    Definition 1 does not imply impossible. Not sensible, yes. Not a good idea, yes. impossible, no.

    as for Definition two, that would be a coloquialism, or at best a regionalism (see the North Ammerican notation) which means that definition is to be avoided in technical and international writings as context can, and probably will, be lost due to the coloquial or regional nature of that definition.

  • (disco) in reply to accalia

    Not disagreeing with what you're saying, but if you're gonna pick up on dictionary usage, it would be useful to be able to spell.

    accalia:
    coloquialism … coloquial
  • (disco) in reply to accalia
    accalia:
    Definition 1 does not imply impossible. Not sensible, yes. Not a good idea, yes. impossible, no.
    [image]

    There are 2 types of "not sensible or realistic". There's the type where someone tells you that and you prove them wrong. And there is the type where someone tells you that and they're right.

  • (disco) in reply to anotherusername
    anotherusername:
    There are 2 types of "not sensible or realistic". There's the type where someone tells you that and you prove them wrong. And there is the type where someone tells you that and they're right.

    and which one is this pray tell?

  • (disco) in reply to accalia

    If you re-read what I said, you'll see that you're the only person who gets to decide that. You get to prove me wrong, and I get to assume it's the latter until you do.

    And you basically claimed that all robots can be built in such a way that it has physical interlocks to prevent it from doing anything dangerous. I feel pretty justified in thinking that you'll never prove me wrong.

  • (disco) in reply to accalia
    accalia:
    Under the laws of common grammarFoghorn Leghorn i was merely, i say, merely being emphatic in my statement.

    FTFY, HTH, HAND.

  • (disco) in reply to anotherusername
    anotherusername:
    And you basically claimed that all robots can be built in such a way that it has physical interlocks to prevent it from doing anything dangerous.

    -_-

    you asked foir a way to prevent the robot from harming itself by attempting to move part of itself through itself.

    now you're trying to twist my argument to be me arguing that physical interlocks, a thing that can be used to prevent a robot from damaging itself, can be used to prevent any dangerous behavior.

    Margelous :moving_goal_post: there Blakey.

    :rolleyes:

  • (disco) in reply to accalia
    accalia:
    Margelous :moving_goal_post: there Blakey.

    Nice job mixing your TV shows.

    He's Blakey, he's Blakey, you're Blakey, I'm Blakey. Are there any other Blakeys I should know about!?

  • (disco) in reply to accalia

    Okay... you still basically claimed that all robots could be built in such a way that they have physical interlocks to prevent any part of it from damaging itself or another part of it... without adding enough size/weight to make the robot no longer capable of performing its intended function(s).

    That's a pretty radical claim.

  • (disco) in reply to anotherusername
    anotherusername:
    That's a pretty radical claim.

    well the physical interlock does not have to witstand the force of the actuator, it could instead physically remove power from the motor.

    granted that does mean it requires manual intervention to unstick but it massively reduces the size of the interlocks.

    so yes i do claim that the physical interlocks can be created to do that.

  • (disco) in reply to FrostCat

    Well there is that thread....

    Filed Under 7826

  • (disco) in reply to anotherusername
    anotherusername:
    "Impractical" and "for all practical purposes, impossible" mean basically the same exact thing. Especially when calling it "impractical" is more of an understatement than it is even the tiniest bit hyperbolic.

    But "impractical" is often used as hyperbole for, "We'd rather not spend that much money." In fact, I think that is the most common use.

  • (disco) in reply to accalia
    accalia:
    so yes i do claim that the physical interlocks can be created to do that.
    I can think of one immediately, in fact.
  • (disco) in reply to FrostCat
    FrostCat:
    He's Blakey, he's Blakey, you're Blakey, I'm Blakey. Are there any other Blakeys I should know about?

    There should be three more, I think?

  • (disco) in reply to Maciejasjmj

    So which one is Avon? Or for that matter, who is Servalan around here?

  • (disco) in reply to accalia
    accalia:
    it could instead physically remove power from the motor

    That's still only half a solution... you'll still have an arm in motion, and possibly carrying thousands of pounds of metal.

  • (disco)

    This thread needs some guidance in objective:

    [image]
  • (disco) in reply to anotherusername
    anotherusername:
    you'll still have an arm in motion,
    If you've got something heavy and inertial powered by an electric motor, you don't simply turn the power off. You use dynamic braking, where the motor itself is the brake. My day job is chairlifts, drag lifts and cable cars - the "normal" brakes are *only* used to hold the system still once it's been braked to a standstill through the motor. The sequence, from a fully running and fully loaded system is:
    1. ramp speed down to zero using the motor
    • engage mechanical brakes
    • remove power from motor windings

    If you do 1 - 3 - 2, you end up with a system that's running out of control, usually backwards. That looks something like this (go to about 4:25 for the money shot):

    https://www.youtube.com/watch?v=u4WPSZojtyE

  • (disco) in reply to Maciejasjmj
    Maciejasjmj:
    FrostCat:
    He's Blakey, he's Blakey, you're Blakey, I'm Blakey. Are there any other Blakeys I should know about?

    There should be three more, I think?

    That wasn't what I was thinking of but it still made me LOL.

  • (disco) in reply to Arantor
    Arantor:
    So which one is Avon? Or for that matter, who is Servalan around here?

    I'm sure he's some sort of sick hybrid of the two, like the guy they stole Slave from wanted to do with the crew.

  • (disco) in reply to loose

    There are a lot of variants on that theme:

    https://itshambles.files.wordpress.com/2013/01/tree-swing.jpg

  • (disco) in reply to Steve_The_Cynic
    Steve_The_Cynic:
    the escaped pieces do all sorts of interesting (for NTSB-related definitions of "interesting") things to aircraft. And sometimes directly to people in those aircraft.
    I was thinking about this last night while flying home. I was sitting where, if this were to occur, I would probably never know it happened.
  • (disco) in reply to tufty
    tufty:
    remove power from motor windings

    What if you do 1 - 3 - 2 but change 3 to

    • reverse the power of the motor windings
  • (disco) in reply to anotherusername
    anotherusername:
    physical interlocks to prevent any part of it from damaging itself or another part of it... without adding enough size/weight to make the robot no longer capable of performing its intended function(s).
    anotherusername:
    The robot may be capable of moving loads weighing many hundreds or thousands of pounds, and all that force is going to go slam against whatever brakes you built to try to stop it.

    A physical interlock need not be a mechanical brake that has to stop hundreds or thousands of pounds. It can be a switch that cuts power to the motor that is moving the robot arm. It's an electrical interlock, but it's still hardware, not software, and it can be tiny (with a big relay to control the big motor, if necessary; still hardware).

    :hanzo:

    Edit to address this:

    tufty:
    If you've got something heavy and inertial powered by an electric motor, you don't simply turn the power off. You use dynamic braking, where the motor itself is the brake.
    However, using the motor itself as the brake for an emergency stop need not (necessarily) involve the normal control system; it can still be a fully hardware thing. For a small-scale example, consider power tools that have a brake to stop the blade when you turn the tool off. A single switch disconnects power to the motor and shorts the motor terminals. The spinning motor acts as a generator/alternator, and the short is a very heavy load on the generator. Kinetic energy -> electrical energy -> heat. That can reduce the kinetic energy to zero pretty quickly. It might not be suitable for every conceivable system, but it covers a lot of them.
  • (disco) in reply to HardwareGeek
    HardwareGeek:
    a big relay

    Which can still fail. Very early computers used relays for a lot of their logic, and their failure rate was much higher than the solid state systems we've used in my lifetime.

  • (disco) in reply to dkf
    dkf:
    Which can still fail.

    Of course. We're talking about a safety mechanism to protect against damage due to software failure or operator error, not something to reduce the failure probability to 0.0000000000000000000000000000000000000000000000000000000000000%; nothing can do that.

    dkf:
    Very early computers used relays for a lot of their logic, and their failure rate was much higher than the solid state systems we've used in my lifetime.
    Not really the same sort of thing, though. A relay that is switching at a very high rate is going to have a higher failure rate that one only switches rarely when an operating limit is exceeded. Also, solid-state relays are a thing, with current ratings up to ~150A readily available. That might not be enough for @anotherusername's scenario of a motor moving many thousands of pounds, but it's adequate for a heck of a lot of uses.
  • (disco) in reply to dse
    dse:
    What if you do 1 - 3 - 2 but change 3 to reverse the power of the motor windings
    That's actually what you've already done in 1, in most cases. There's two types of motor braking used on chairlifts. The most common one involves reversing the phases and ramping the current against the speed, which brakes the lift against the torque of the motor, giving a smooth curve down to zero. The other uses pulsed braking, which is more efficient in terms of power used, but far less smooth and far more susceptible to rollback events.

    The reason you don't do 1 - 3 - 2 is that you end up with, between 3 and 2, an unbraked system which is usually heavily loaded on one side. The emergency brake should catch that situation pretty much as soon as the lift starts rolling back, but even a 50cm rollback is noticeable and scary (for the passengers as well as the driver). I've seen a 2m rollback on a "live" lift before, and I don't want to see it again.

    Some lifts have mechanical antirollback. Effective, but brutal.

    HardwareGeek:
    However, using the motor itself as the brake for an emergency stop need not (necessarily) involve the normal control system; it can still be a fully hardware thing.
    Emergency stop is, at least for chairlifts, an entirely separate system, decoupled from the motor. You have a big fuckoff brake on the bullwheel. Discordance between motor speed and cable speed (might be a broken gearbox)? Brake slams shut. Lift stops. Cable catcher security break? Brake slams shut. Lift stops. Rollback? Brake slams shut. Lift stops. You get the idea.
  • (disco) in reply to HardwareGeek
    HardwareGeek:
    A relay that is switching at a very high rate is going to have a higher failure rate that one only switches rarely when an operating limit is exceeded.

    I'd be more worried that the system would fail-closed and nobody would notice.

    (I did like the shorting of the motor terminals approach. A nice, physical “stop now means STOP NOW!”)

  • (disco) in reply to accalia
    accalia:
    well the physical interlock does not have to witstand the force of the actuator, it could instead physically remove power from the motor.

    Let's say that in that situation the arm in question is moving a 5 ton load when you cut the power. What's stopping the momentum of the load after the power is cut? Your fail safe just failed.

  • (disco) in reply to tufty
    tufty:
    If you do 1 - 3 - 2, you end up with a system that's running out of control, usually backwards. That looks something like this (go to about 4:25 for the money shot):
    That whole sequence from about 4:45 to 6:something (depending on whether you want to watch the aftermath shots) was pretty golden. Well, not if you were one of the guys standing there, I suppose. I'd have said the money shot was at about 5:27 with the hard-hat guys running away.

    bang chair destroyed bang chair destroyed bang chair destroyed bang chair destroyed Hey guys, where are you going? I can keep this up all day...

  • (disco) in reply to PWolff
    PWolff:
    (Btw, vae is Latin for woe)

    va'e is lojban for "nth position on a scale".

  • (disco) in reply to ben_lubar

    Here's the idiomatic Esperanto for Santayana's famous quote

    Those who cannot remember the past are condemned to repeat it.

    [spoiler]Hej , ni lernas lobjan![/spoiler]

  • (disco) in reply to tufty
    tufty:
    [spoiler]Hej , ni lernas lobjan-on![/spoiler]
    FTFY
  • (disco) in reply to Scarlet_Manuka
    Scarlet_Manuka:
    *bang* chair destroyed *bang* chair destroyed *bang* chair destroyed *bang* chair destroyed Hey guys, where are you going? I can keep this up all day...

    It couldn't, though... mangled chairs getting stuck in the mechanism was what eventually stopped it.

  • (disco) in reply to Scarlet_Manuka
    Scarlet_Manuka:
    I'd have said the money shot was at about 5:27 with the hard-hat guys running away.

    I don't know why they stood there that long. I'd have run or taken cover behind something strong and heavy much earlier. Maybe the perspective on the shot is misleading though and they were further away than they look.

  • mb8 (unregistered)

    The real WTF here is calling clamp when comparisons are already done in the assertion...

  • JamesPaync (unregistered)

    XRumer is the multipurpose tool for comprehensive advance on the Internet

    The program is capable in a short time to place announcements and your references more than on 400.000 resources — forums, blogs, guest books, reference catalogs, bulletin boards, social networks . XRumer bypasses also text protection against robots — questions like "How many there will be 2+2?", "Call the capital of Russia" and so forth. The program knows more than 170.000 answers to similar questions, arithmetics all is counted automatically , the knowledge base is replenished every month . Ample opportunities advance in social networks
    It is free the enclosed SocPlugin allows at once in 4 social nets — Odnoklassniki, VKontakte, Mamba and Facebook — to distribute messages, comments to video and a photo , to load video and a photo , to distribute invitations in groups and friends — at the same time besides on the full automatic machine bypassing everything probable protection against boats and mailings, including kapch. Set unique developments allows to try to obtain very much good and interesting results Such hi-tech functions, as set thematic mailing "Antispam" (in any the forum is sent the message according to scope of this forum, irrespective of the region and language ), a system "Question-answer" and others — allocate XRumer among others similar complexes, creating opportunities it is valid "clever" and also socially useful mailings. In long-term opportunities XRumer — successful investment If you the owner online store or Internet service, for certain understand as advertizing in Yandex-Directe or Google AdWords. Competently applying the XRumer + Hrefer + SocPlugin complex, you receive a possibility of everything for $10 a month get necessary to you target visitors .

    Minimum technical requirements Windows Vista/7/8, 8.1/10 (both 32 bits, and 64 bits); Windows Server 2012 R2 (but Win Server 2016 and 2019 NOT are recommended!); The proce

Leave a comment on “A Hardware Switch”

Log In or post as a guest

Replying to comment #:

Log Out

« Return to Article