h a l f b a k e r yWhy not imagine it in a way that works?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
In many industrial applications, it is necessary to lock out power to a piece of equipment before servicing it. Shutting off the power isn't good enough; the power switch must be locked "off" using a padlock whose key is held by the person working on the equipment.
In many environmental situations,
this is handled by having a "lockout board" from which hang a number of color-coded locks with their keys installed. A worker who needs to lock out a piece of equipment will take a lock, put it on the device in question, and take the key. When he's done, he'll remove the lock and return it to the board.
While this approach is certainly workable, it does have two disadvantages:
-1- A worker must keep the keys for all lockouts he controls.
-2- Inspection of an installed lock will give no clue who holds the key.
What I'd propose would be a padlock which is designed so that it can auto-adapt to any of a few dozen keys. When not in use, the lock would be open. Inserting a key, closing the lock, and removing the key would "set" the lock to use that key; unlocking it would require reinserting the key. As an additional feature, the lock could incorporate a readout showing information about the identity of the key that was used to secure the padlock.
One possible implementation would be to use a cross between a settable combination lock and a tumbler lock. Inserting a key would 'dial the combination' on the interior mechanism, and opening the shackle would allow the combination to be reset. To allow easy identification of the employee responsible for securing the lock, one or two wheels/sliders would be exposed to a window allowing that part of the combination to be seen.
[link]
|
| |
I see what you're getting at here. |
|
| |
However, I'd point out that any situation with which I am familiar has the worker getting assigned his/her own (set of) lock-out locks and keys, the locks being engraved with the worker's name and the corresponding key number. I believe this effectively solves the problem illustrated in your #2 point. |
|
| |
As far as your #1 point goes, that's an inherent aspect of keys/locks, it seems to me. |
|
| |
This is certainly workable with an electronic lock. If I enter my (secret) number on the keypad, the lock closes and a panel displays my employee ID. |
|
| |
In more sensetive applications, the whole lock/key issuing system is tied in with a permit-to-work documentation trail as well - it's laborious but does assure a high level of system integrity. We'd be very loath to advocate any change that might move away from this type of highly supervised system. But we see the point, and it's a good one for lower-risk applications. Sorry, but we can't quite bring ourselves to give it a croissant yet .... |
|
| |
angel: even easier if you use fingerprint recognition. Though perhaps hands get too dirty in an industrial situation...? |
|
| |
Wouldn't know. Long time since I was in an industrial situation. |
|
| |
Why not just get the workers to sign for the locks and keys when they take them from the communal cupboard? |
|
| |
Well, one guy may need to have several systems locked out at the same time. |
|
| |
I like the idea of the electronic padlock. Rather than typing a code on a keypad, I'd prefer a card swipe. You could lock as many padlocks with one card as you like. Plus, you could lock one padlock with as many cards as you like. That way, if Bud and Bubba are both working on the power line, but in different locations, each can swipe his card through the lock. The lock remains closed until both Bud and Bubba swipe their cards though a second time. An LCD can show the number of outstanding lock requests and the ID number of the most recent requestor. This feature would be very easy to implement in software. |
|
| |
Until a hardware solution exists to track lockout owners automatically, I guess people will just have to resort to procedural methods of tracking. Either have an info tag accompany the lock on the equipment, or else have a lock checkout system with a log book of who is currently using which locks. Both methods are in wide use at various locations. |
|
| |
The difficulty with having locks pre-assigned to users is that it becomes necessary to issue each user the maximum number of locks they'll ever need. In some cases, a user may very seldom need more than two or three, but in one particular case might need a dozen. If multiple users wouldn't need a dozen locks all simultaneously, the communal lock approach might be good. |
|
| |
Actually, if safety regs would allow for electronic locks in these applications, they could have substantial advantages over conventional ones. For example, if a worker installs a conventional padlock in a breaker-panel lockout spot without using a multi-lock adapter, it may be impossible for someone else to attach an additional lockout. An electronic lock, by contrast, could keep track of multiple active lockouts. [My mechanical lock could be adapted for that also, though operation would be a bit less intuitive and the mechanism would increase in complexity with the number of lockouts.] An electronic lock could also provide logging abilities. |
|
| |
Include in the padlock a hole through which the shackle of another padlock may be placed. When another padlock is inserted in this hole, the lock may be unlocked but will not actually open until the other padlock is removed. |
|
| |
Sorry, but we have to advise against anything involving card swipes or digital codes .... NEVER USE SOFTWARE IN A SAFETY CRITICAL APPLICATION. By which we mean detonator systems, HV switching lockouts, and anything involving nuclear material. |
|
| |
It may seem stupid but when you've had experience of this stuff, you'll be reasured when you work by the knowledge that the control system is blocked by a piece of 8mm hardened steel bar, and the only key is on a string round your neck. Safety regs specifically exclude the use of electronic systems for a number of very gooAbort, Retry, Ignore ? |
|
| |
Current drive is no longer valid> |
|
| |
Unable to find COMMAND, system halted. |
|
| |
General Protection Fault ........... |
|
| |
Unless, of course, the software is necessary for responses that are faster than a human can react, such as fly-by-wire systems, flight directors, atmospheric re-entry systems and anti-skid controls. |
|
| |
The point is that there are going to be more and more situations where it will be considered unsafe for a human to have control and only software will be able to fulfill the role. |
|
| |
fly-by-wire => Die-by-wire. No thanks. |
|
| |
But actually, this isn't about control; it's about disconnecting control (Like cutting off HAL's cabling in 2010). It's one thing to have the software in control most of the time BUT you still need a manual override/hard interlock. It's not about speed of response - it's about ensuring no response at all. |
|
| |
We've seen software, and we've seen clunky stuff with big levers and wires and cogs, and when you're dealing with the real Scary Stuff the levers and cogs are definitely better. |
|
| |
Soory it already exists. CHeck out Cyberlock by Videx |
|
| |
too complicated. In industrial situations a combination of locks and safety interlock devices are used. Here is how i believe it works: Each primary carries a key with their name on it. Any one of the primaries can shut down the circuit. The circuit can also be shut down using an emergency relay which can only be reset by inserting the supervisor key. When a key is inserted a interlock bar pops out of the breaker locking the key in place, indicating who disabled the circuit Whomever is doing the service takes the unique interlock bar with them and returns the bar when the service is finished, returning the bar allows them to retrieve their key. With out the uniquely shaped bar the circuit cannot be energized. |
|
| |
In most lockouts, I believe you're supposed to write your name on a tag that is attached to the lock. |
|
| |
An LCD display or mechanical numbers that dial up could provide this information automatically. |
|
| |
Another idea I thought of was to have a lock with 2 key holes. Every employee has their own key with their name and phone number on it. When the lock is not in use, a key is kept in the 2nd keyhole and can't be removed until the employee puts his/her personal key into the lock. Then lock the lock, and turn the common key in the 2nd hole, to the removal position and take it. This locks the employee's key into place. |
|
| |
For multiple employee lockouts, a solution exists for that already-it looks like a handcuff with a pegboard attached to it. You put the handcuff part through the hole of whatever needs to be "locked out" and close it. Then you can put up to (usually) 6 locks through the holes of that thing. |
|
| |
Usually use a key exchange built into door locks. The exchange has 4 or so keys that have to be returned before the door can be locked, main key removed and then inserted into whatever it is that turns stuff on. Only one of each key exists and anyone loosing one gets shouted at. |
|
| |
Totally agree with 8th - keep computers out of safety systems if at all possible. Quite prefer relays, pneumatics and huge bolts. |
|
| |
// Inspection of an installed lock will give no clue who
holds the key. // |
|
| |
In every industrial environment in which I have ever
worked, anyone who locks out a device is required to put
thier tag on the lock. These tags have names, company
ID#s, sometimes photographs, and occasionally a dry-erase
section for the date the device was locked out. Failing to
use your tag/remove your lock when the job is over can
land you in hot water. I have never been to a place where
lockout procedures are used that did not issue lockout
tags. Point #2 is moot to my eyes. |
|
| |
As for all other points, I agree with (and I can't believe I'm
saying this again) [8th]. Simpler is safer. |
|
| |
While "post-it" notes have a pretty feeble stickiness to them, allowing too-easy peeling-off, they do make stickers that stick tighter and take some effort to peel off (but they do peel cleanly). |
|
| |
So, I propose getting a batch of such stickers and when an empolyee takes the lock, put a sticker on it and write his/her name on it. Now you know who put the lock on the machine. If the tags are big enough, multiple names might be written on it (allows work on a "down" machine to continue past a shift-change). |
|
| |
Regarding multiple keys, a Policy change might be able to deal with that. If all the locks used the same key, then each relevant employee only needs one key. If Employee A opens a lock that Employee B name-tagged, and Employee A was not one who worked on the machine that was locked out, then Employee A is fired, period. |
|
| |
That's why each lock uses a different key. Policies like that
don't always hold up well in industrial workplaces. |
|
| |
// Employee A is fired // |
|
| |
... which will be a consolation to Employee B's family at the Memorial Service. |
|
| |
[8th of 7], in that case Employee A can also be arrested for first-degree murder, and later executed. Informing all employees in advance that those are the consequences should lead to additional incentive to avoid removing locks that they weren't involved with installing. |
|
| |
Again: different keys, no confusing procedure policies, no
murders. |
|
| |
The other point is that contractors like myself use our own
locks (and our own lockout tags, usually saved from old
jobs with a new sticker pasted on for the date), which
frequently means your going to have a lock on there with a
different key anyway. |
|
| |