Very interesting lock

[webidiot2]

Found this in google patents.

Rotary pin-in-maze discriminator

I don't believe I have ever seen a maze lock before. I guess you could say I'm A-Mazed by it lol.

[Legion303]

That looks amazingly easy to work out...

-steve

[Dak]

it does look easy to work out, but i must admit that IS creative

[ridinplugspinnaz]

That looks amazingly easy to work out...

-steve

If they had invented some sort of modification to the design that wouldn't give the operator such obvious feedback as to when they've accessed the incorrect state, that could end up being a very interesting device... Even still, who wants to a 24-character binary passcode?

[ridinplugspinnaz]

Don't get me wrong though, that's a really nice find, webidiot. The more I think about it, the more I think that would actually be a somewhat viable lock design if they could do something like shunt the "guide finger" (for lack of a better term) off onto an ambiguous decoy track once the wrong gate is chosen in the sequence... maybe something in the third dimension, like ramp the guide finger up onto a track that won't open the lock, but still has gates for the operator to dial through as if they were on the right track.

I guess the biggest downside to that sort of modification is that you have to dial in 25 "characters" before you realize that you entered the combination incorrectly... assuming you can remember that you've counted that high!

In fact, thinking about it a bit further, some of the methodology in this patent has a fairly direct mapping to computer science analogues, particularly the choice of binary states for each gate of the locking mechanism. Maybe it's the cynic in me, but I have to wonder if someone is patenting a physical mechanism of arguably dubious utility just to get some prior art out there for things they're doing with computers? It looks like this patent came out of Sandia (as in the national labs), so I wouldn't exactly be surprised. Heck, why not encode your ideas for algorithms in a physical patent these days? At the rate software patents are being reversed and obviated, it might not be too bad of an idea...

[webidiot2]

I wish you guys had read the abstract or the description before jumping to conclusions. If the lock is given the incorrect signal twice the lock seizes up and cannot be unlocked. Although not unpickable but would imagine very difficult.

[Major Boothroyd]

What about as a combination of a digital key and a mechanical lock? Automate the maze in accordance to the code provided by a key card or something like it? Probably more complicated than it's worth and not much better or maybe even worse than current digital locks, but thought I'd share the first thing that popped into my head when it required you to dial 25 characters, especially since I've forgotten my 4 digit bank pin before.

[ridinplugspinnaz]

I wish you guys had read the abstract or the description before jumping to conclusions. If the lock is given the incorrect signal twice the lock seizes up and cannot be unlocked. Although not unpickable but would imagine very difficult.

To be fair, I went back and speed-read the entire patent, and couldn't find any language that specifically indicated that upon a second incorrect attempt, the device becomes irretrievably unlocked. Can you cite that part? Perhaps I just missed it, but I also don't appreciate the insinuation that none of us read anything about it before commenting.

In reality, using this device as a lock to prevent unauthorized access, versus using it as a safety interlock, implies two very different desired outcomes w.r.t operating characteristics. As a safety interlock, you most certainly want feedback from the device when the wrong code is entered into the discriminator -- you want to know when something went wrong. If using this lock to keep others out, you probably want to minimize the feedback that the lock offers to an unauthorized operator so that the lock is more difficult to manipulate.

Just my $.02.

[ridinplugspinnaz]

^^ I meant to say "irretrievably locked"... sorry for the confusion.