My unpickable lock design

[Daggers]

So i decided that im going to reveal my design i've been working on for a totally unpickable lock. So imagine a regular pin tumbler lock. My lock has a sidebar coming in from the side that the edge is bevelled so that when the sidebar goes into the lock, the angled edge lifts the pins to sit on top of the sidebar. This position is how the sidebar is normally, it sits inside the plug completely and has the pins on top of it. Now, infront of the pins on the flat part of the sidebar, it protrudes upwards to the height of the keyway to block any view of the pins. The front and back of the sidebar is angled as is the tip of the key to allow the key to push the sidebar out. When the key is fully in, the sidebar slide back into the keyway through a hole in the key to match the sidebar. Now the plug can turn. If the attacker tries to gain access to the pins by moving the sidebar out, the tolerances will be so that the sidebar binds before the pins (just make the pins a little looser in the chambers than the sidebar). The pins will go up but won't stay up or bind at all. In order for the pins to bind, the sidebar has to be in, but then the pins are not accessible. In order for the pins to be accessible, the sidebar has to be out but the pins can't bind. It's a circle with the key as the only solution.

This is a horrible explanation of the lock still, but i can't upload the attatchments since they are in e-drawings. you can view the e-drawing files by downloading the e-drawings free viewer here: http://www.solidworks.com/sw/support/downloads.htm and if you want the files, send me a message with your email and i will email them to you.

[vov35]

makes sense, however I feel such a design could be decoded. Clealry, it has no resistance other than the proprietary nature of the blank to traditional impressioning.

Also, the way you described it, no pin position can be lower than that of the sidebar, which means that a sufficiently thin pick can be inserted. It would certainly be difficult to open, but I'm sure some members of this forum would have a field day if you labeled it "high security".

[cybergibbons]

No lock is unpickable. The act of picking is simply that of emulating a key. If the lock can be opened with a key, it can be picked.

[jonwil]

I genuinely believe that NO mechanical lock can ever be truly secure against a determined attacker with the right tools and knowledge and that a high security electronic lock (i.e. NOT the pieces of junk one finds at most big box hardware stores) could well be the best security.

Done RIGHT (and without the security flaws that have been found in a number of hotel locks as of late) I believe it is possible to produce an electronic lock which would be essentially impossible to break without either access to a working key or using destructive entry of some sort (and there are ways to make destructive entry much harder too). And if you use the right hardware, I believe you could produce a key that is pretty much impossible to copy without destroying the original.

[Lock Jock]

And if you use the right hardware, I believe you could produce a key that is pretty much impossible to copy without destroying the original.

The right hardware in this case would have to contain RAM (or EEPROM, whatever your pleasure) on-die within the cpu itself, since cpu>mem bus dies have been hacked (with difficulty and expensive tools, but nothing beyond the capabilities of spies, both military and corporate).
You may have heard about this old story: http://www.usatoday.com/tech/news/compu ... cked_N.htm

[jonwil]

Yes its possible to hack these chips but it requires specialist tools and knowledge. And it destroys the chip (or at the very least makes it obvious that someone has tampered with it) which means that (unlike pretty much every mechanical key I know of) its not possible to copy the key without someone knowing about it.

Which means that if a key is given to someone temporarily (e.g. cleaner, tenant, employee, whatever) it will be immediately obvious that they have tampered with it (and hence you can revoke that key from the lock)

To prevent someone from extracting the data from the key and making 2 new keys (one of which is then returned to its owner as the "original") you can attach special nearly-impossible-to-copy labels to the case and chip (similar to the "warranty void if removed" stickers they use to cover screw holes on electronics)

[FarmerFreak]

you can attach special nearly-impossible-to-copy labels to the case and chip (similar to the "warranty void if removed" stickers they use to cover screw holes on electronics)

FYI, there are tamper evident competitions run at hacker conventions. Those "nearly-impossible-to-copy labels" aren't nearly as nearly as impossible to copy as you may think.

[raimundo]

I forget the circumstances but someone needed a key that can't be copied by the kid at the cole national booth or whatever they have nowdays, so I told him about using a flat file on the bottom edge of a key angled to cut more toward the tip so that it cant be put in the jaws of the duplicator effectively, but since that is not something I have ever done, its only an idea somewhere in the deep pile that is lp101
It would seem to probably make a key that has to be jiggled, since its riding the wards rather than the bottom of the cylinder,
and what happens when you turn to 180 degrees, do top pins and thin master wafers drop in at the tip of the key?

[Lock Jock]

you can attach special nearly-impossible-to-copy labels to the case and chip (similar to the "warranty void if removed" stickers they use to cover screw holes on electronics)

FYI, there are tamper evident competitions run at hacker conventions. Those "nearly-impossible-to-copy labels" aren't nearly as nearly as impossible to copy as you may think.

Nor are they nearly impossible to remove intact and reattach with the owner never realizing it. As a consumer electronics tech, I have a spare roll of warranty-voiding adhesive labels for my repairs which are exactly the type to which jonwil refers. I rarely used them, leaving a great many on the roll for practice.

I've also yet to meet a car registration or inspection sticker that I couldn't remove intact -- those are easy, though, so it's not really a fair comparison. No, I'm not a label (or car) thief.

Point being, there's no need to copy such a label if you can remove it, do your dirty work and then successfully reattach it.

One could try the translucent, tamper evident paint (the kind factories use on alignment screw heads, et al) but this chemical can be purchased by the public if one knows where to look -- or produced if one is a chemistry wiz.
Even as an electronics tech who dealt with this stuff on a daily basis, it took a good deal of time and effort to discover the trade names of this product and even longer track down a supplier.
No, it ain't Glyptal! Some manu's will try to tell you that's what they use; it isn't.
No, it isn't any Loctite formula, including the aerobic ones.
And no, it certainly isn't the poor (or lazy) man's version: nail polish.
Electronics manufacturers will not sell the stuff and the real McCoy is not something you can get at Home Depot or just about anywhere else you've heard of.
Point being, since I found it, other motivated persons probably have or will.

A much better solution than any external tamper evident scheme would be internal booby traps which result in code erasure. This obviously requires the owner to have noted the code and kept it in a safe place for reprogramming of the original key (if recovered and still functional) or a new key.
I've been developing some of these traps for electronic devices, including keys, but as I've not found them in patent searches, I'm clamming up about that for now.

[cybergibbons]

I think you need to check the meaning of impossible. It means "not possible" not just "really hard".

[ice_man]

its sounds good but no lock can be unpickable its just makes it harder

[jonwil]

And if you use the right hardware, I believe you could produce a key that is pretty much impossible to copy without destroying the original.

The right hardware in this case would have to contain RAM (or EEPROM, whatever your pleasure) on-die within the cpu itself, since cpu>mem bus dies have been hacked (with difficulty and expensive tools, but nothing beyond the capabilities of spies, both military and corporate).
You may have heard about this old story: http://www.usatoday.com/tech/news/compu ... cked_N.htm

Yes chips can be decapped and read, even "high security" ones.
But once you decap it, the chip is permanently damaged and even though you may be able to get the data back, you cant get the chip itself back into a state where it still works AND shows no evidence that it was decapped (and this applies to even the best-equipped money-is-no-object government spy labs).

If you can show me a way to decap one of the higher-security microcontrolers (with ram and rom on-chip with the CPU), read its contents and then restore the key back to original condition so its not possible to identify that its been tampered with, feel free to enlighten me. And no burning the key contents to another chip isn't an option if you use the right part (there are MCUs with unique serial numbers burned directly into the silicon, good luck replacing that one with a new one in a way that cant be detected)