@@Yuki2204 saw one of these used at the bank I worked at when a teller screwed up a combination change. We still used it for years afterwards with no ill effects.
@@ottokiehl5413 1337 means "leet" as in "elite". It became a kind of language where letters would be replaced with similar looking numbers if applicable. for example, "Hello" would be like "H3110" big joke in gaming communities usually
Thieves don't usually use this type of machine - it's main use is situations where the legitimate owner of the safe would like it opened non destructively but forgot the combination and an expert safecracker isn't available.
@@zerog2000 Trinamic stepper drivers are pretty good. They are very silent and also come with other cool features like detecting stall by measuring back EMF, also with load-dependent current control (motors will run cooler) and much more. I wonder how much that robot cost.. it looks like it is some kind of prototype (3d printed parts)
I expect a real servo motor would be superior. They are faster, quieter and can be more accurately programmed. The downside is that they are more expensive.
Just imagine walking through your office at night and hearing what sounds like a 3D printer, going to the safe room and finding this thing whirring away on its own
I don’t know how you manage to get so many people, myself included, to watch you pick locks almost every day. Not complaining, it’s truly fascinating to see your raw skill.
@@VectorxMan he did not even elaborate to make clear what is that he made to improve or whether he built the thing from scratch. If it was me I would have bragged about it for two hours
I reckon you could speed up the algorithm by skipping the "long distance" numbers, its much faster doing "close by" numbers, so if you prioritise testing those combinations first you get through a larger portion of your testing regime sooner.
That won't work because of how it is testing. As you see it is holding numbers one and two until it tests all of number three. It does this because if it wants to change numbers one or two it has to go all the way around the dial in reverse. Good thinking though.
I thought the same thing. Essentially keeping the total of the 3 lock combination numbers as low as possible for as long as possible. Letting the cracker go through more numbers at the beginning and slowing down as it goes through the combinations.
For those wondering how it knows when it hits the right combination: After the last digit is correctly guessed, the dial will only turn back so far before it stops spinning. The robot monitors how much resistance there is and when it's suddenly prevented from spinning the dial, it knows it hit the right combination. If you go to 10:50 and watch in 0.25x speed you can see what's going on pretty clearly. As the middle gold disk turns clockwise, this is the robot entering the next digit (notice it spins a bit further each time), and at the end of each clockwise movement you'll see the outside ring with the gate in it move one more tick. It then goes counter-clockwise back to the previous position before trying the next digit. However when all digits have been entered correctly and the gates are all aligned, and the dial starts to turn counter-clockwise again, it allows the hook to fall into place (the loud click in the video) and stop the dial from turning back to the previous position. Since I'm this far, I'll include how the hook & rest of the physical lock works: If you look at the top right of the hook there's the white indent looking bit, that's a piece that's deeper than the rest of the hook that needs all 3 gates aligned so that the hook can move to the left and catch on the cut out bit of the gold ring (when the gates aren't aligned, that bit is just resting against the outside of the 3 rings). At that point, as you(/the robot) continues to spin the dial counter-clockwise it pulls the hook up, including the gold deadbolt looking piece that the hook is attached to. That deadbolt bit is what's physically blocking the silver cylinder at the bottom, which is connected to the bolts keeping the door closed. You can see at 10:56 when he retracts the bolts that the cylinder goes where the deadbolt originally was. This is why when you close a safe after you lock the bolts (normally turn the handle back to the closed position, if you're using the safe normally) you need to spin the dial to properly lock it. It pushes the hook out, which moves the deadbolt back down, and spins the gates so they aren't aligned anymore.
This ia a nice directive explanation 👍, thanks! I'm surprised how little clearance is involved in keeping the safe bolt blocked. The fact that this explanation only got 23 likes, 1 from me is very telling about the audience as a whole and a little sad. Hence LPL gets paid, they don't.
I own a small CMI fire-rated safe with this SG combination-lock fitted, but it won't open despite using what I'm certain is the correct combination as dial does seems to stop where it should when using those numbers. Can these mechanisms bind if left to gather a bit of dust & grime? Story is that I purchased safe in mid 1995 and was using in my jewellery manufacturing business before retiring due to ill-health. After sitting unlocked in my garage for about 5yrs, a friend shut & locked it before I could stop him as I wanted to check it, & now 5yrs on I'm still trying to open it as I'm in the process of moving and I would like to use it at my new residence.
@@stevie-ray2020 Binding would be a pretty serious failure. If it was common there would be complaints online. Stating the obvious just in case: 1) the S&G knowledgebase articles e.g. stuck lever and misaligned wheel 2) the basics: mis-remembered combination order / turning the dial the wrong way / forgot you changed combination / etc. Good luck! You can always buy a robot, drill and repair it, learn lock manipulation or take it to a locksport convention as a challenge.
When he says people wanted to see the robot run for longer I thought "who would want to watch a robot run for 9 minutes?" I then proceeded to enjoy watching a robot run for 9 minutes.
I mean, a minute or two sure, just to get an idea of how it runs at normal speed, but it would be interesting to see how the mechanism is turning the dial in slow motion. I've always thought that it you mess up one number in a combo safe that you have to spin it and start over from the beginning, but apparently I need to do more research into that. It seems as if there's a shortcut to try multiple secondary and tertiary numbers without necessarily needing to start all the way over.
Hmm I would guess that a simple way in countering this attack is adding a disc with a centrifugal detent. On normal operation one does not spin the dial that fast.
@@eristheslayer hm it’s not really locking you out. It’s just stopping the rotation momentarily, so it WOULD fool the automatic mechanism that detects the opening of the lock. And even if you make it avoid it you still make the dialer slow enough to be impractical
The digital versions the military use have those type of functions: It knows a human can only turn the hand so far, and it requires a variable speed of a human. otherwise it resets. (as far as I know, there is no known surreptitious way to enter these versions.)
@@annakelly9850 Would it simply be possible to read a value from a random number generator for the speed, add ramp up and ramp down features, and simply limit it to the speed/range of a human? Jitter could possibly be be introduced randomly too.
@@Jeff-ss6qt Nope. If that dial spins fast enough it stops the whole process. So, you have to make the dial turn so slow that it'll take a long, long time. You can still use a robot, but the robot won't speed it up, just make it automated.
It was genuinely one of the coolest things I've seen in a long time, I have never seen the mechanism of a combination lock like that work internally and it's super interesting!
10:30 is the moment you want; it cycles back through the previously found spots and then cycles through one more time to unlock 👍 great mechanism and computing
It's the first time that "looks like WE got this open" makes sense. Up until now, he was the only one opening the locks. Now they are two and BOTH are machines.
Fantastic mate. 50+ yrs ago I was a watch and clock repairer in Adelaide. Had a customer with an S&G lock on his floor safe. Fortunately the door was off when the lock stopped working. One of the springs had broken (rusty). It was many years old. I had a LOT of fun repairing it and and repairs to the number change system. Pity so many new safes gone electronic now.
Because he knew the combo ahead of time and he could see it was about to open it. If you have the back of the lock open like he shows, you can get the combo in a minute by turning the dial and watching the wheels align.
I started to watch and then got hypnotized and closed my eyes. I opened up (feeling refreshed, but remembering nothing about it), about 1 or 2 before he unlocked it.
@@ipick4fun27 it looks like "leet". if you google "1337" or "leet' or better yet "1337 leet" youll find all the information you could ask for. theres even a wikipedia page that I didnt know about LOL
@@ipick4fun27 It's supposed to be called "leet" speak like as in the word elite. if you remember back in middle school everybody would write "800813" because it "looks like boobies upside down". same concept with 'elite speak'. You basically just replace letters with numbers wh3r3 4ppl1c4ble lik3 th1s and say things like g3t 0wn3d k1d. It's super cringey even when it became a fad and it thankfully died down fast but you'll occasionally see 'skript kiddies' still talking like that because they think they're an 3l1t3 h4x0r when in reality they barely learned the basics of cheat engine. You can usually spot one from miles away because they'll undoubtable have the F4z3 C14n l0g0. God that physically hurt to type and press enter.
This video is just long enough that the final "click" at the end was super satisfying. Anyone who's ever successfully run a password brute force tool knows how good this feels, when it finally finishes! 😅
Not only in physical locks but also accounts and WiFi password cracking brute forcing is practically impossible especially the security that they implemented but changing the password like 12345678 and the computer says the password is been crack it's a satisfying moment
I kind of know this? Though in a completely different context. The Resident Evil 2 remake that was released back in 2019 has a couple of three wheel six letter combination locks, and I brute forced all of them. Not quite the same, but it still felt satisfying.
As cool as the safe cracker is, I would love to see a video with the saves themselves as the main feature. Either way I love your videos and the way you explain topics with genuine enthusiasm and interest. It shows in your work and is not lost 👌
"Nothing on one, nothing on two ... nothing on seventy-seven, nothing on seventy-eight ..." Auto-dialers are legit tools for locksmiths but it's not recommended that they be used repeatedly on equipment that one cares about. The interior shot should make it plain that they can cause wear of several components.
I can't imagine a situation you would need to use this on the same lock repeatedly. If you are stealing, you just need to use it once, if you forgot the code or just don't know it, you would also need to use it only once. If you have brain damage and are always using the machine to open your safe instead of using the code, then you might have a problem.
@@mindfortress105 Yes, that's my point. It's a legitimate tool but it would be unwise to use it for fun. Doing so might be a sign of brain damage. Or thoughtless, anyway.
if these robot beasties ever became a problem, lock makers could do what IT departments have been doing for 50 years -- too many failures in a row would cause a lockout for a few hours.
You’re describing a use case that is ridiculously rare. Don’t use that poor little lock for fun! If it’s their lock it can be used in whatever way they see fit to use it. And you’re calling others brain damaged. Or trolling
If you’re using a robot to brute force a safe open do you really care about the well-being of the lock? If you cared you’d probably have the combination.
Crap, now I gotta go watch it for the wingnut ? Seen these operate enough back when I owned one. LOTSA of safes it will NOTY work upon, for reasons I will not enumerate.
@@citylockapolytechnikeyllcc7936 Cam here to see if anybody noticed this... :P You can see it clearly if you move the fwd. But it's moving left, so either it is tightening or it is a left handed screw.
With the introduction about 1 minute, 15 seconds to unlock the safe, another 15 seconds to show it's not a fluke, then the rest of the vid showing his comments about how good or bad the lock is.
If I were making a safe, I'd add a mechanical countermeasure : the rotations wind up a spring that powers an escapement mechanism. If the spring gets fully wound-up, say after 500 rotations, the dial gets disconnected (clutch ?) and only gets reengaged to the lock after the escapement stops moving, which you could set to be 24 hours. This would limit severely your attempts per day, and then you can throw your robot in the trash :-D
To be fair, not many criminals can afford money for this thing and time to wait for it to solve the lock. Straight up destroying the safe would probably be faster :p
There is probably an algorithm it could use to calculate travel distance and do shortest travel distance numbers first. It could knock over a bunch of the solution space faster
There are also general rules about the relationships between the digits you’re recommended to not violate as those settings are apt to bind up the lock and such. From S&G’s manual they suggest the following which also would constrain the solution space a lot… 1. Do not select numbers which all end in “0” and “5.” An example of a poor combination is: “50 - 15 - 30.” An example of a good combination is: “52 - 15 - 37.” 2. Do not select numbers in an ascending or descending sequence. An example of a poor combination is: “22 - 41 - 68.” An example of a good combination is: “22 - 68 - 41.” 3. Leave at least a ten number margin between adjacent numbers of the combination. An example of a poor combination is: “41 - 39 - 66.” An example of a good combination is: “41 - 30 - 66.” 4. Do not use a number between 95 and 99 or between 0 and 20 for the third (or last) number of the combination. This can cause a condition which will inhibit the mechanism from locking correctly. 5. Avoid using easily guessed numbers (birth dates, house and phone numbers, etc.).
@@JamesReedy Note that LPL said that certain combinations are omitted, which is why it has ~800,000 possible settings instead of 100*100*100 = 1,000,000.
That can be defeated or delayed if there is a thermal fuse type thing that locks up the mechanism after friction from high speed manipulation heats it up
If the metal gets that hot sure. But metal on metal doesn't make nearly as much heat as your thinking. Then you have the fact a mildly warm day may be enough to equal the heat from this type of attack. You've made it harder to break in, but you've also made the safe completely unsellable to any place that has heat problems
@@G58 unless you were continuously spraying it, something like wd40 is too light and would get wiped away pretty quick with only one coat. Basically, the lock would need some sort of oil bath to handle that kind of abuse without tearing itself apart.
TheOldPioneer Great points. Thank you. It would be interesting to see how much or how little and how frequently dosing would solve the problem, and how well an automated spray feed could be implemented. I find that GG58 leaves a water resistant coating which also lubricates. Maybe a top down delivery squirt every 60 seconds could sort it. The most valuable component in any engine is the oil. It’s also one of the cheapest, easiest to change, and most often neglected. My ‘02 diesel Mondeo has done 200,000 miles, and to be fair I don’t service it according to the manual. I haven’t touched the injectors. But the emissions are .52 where the max is 1.50! Regular oil changes using quality oil. 🧐😎
Blows my mind how fast that thing moves. I'm sure it'll go faster, wonder if it get hot. It was absolutely mesmorizing to watch. I enjoyed it thankyou LPL
Doc and Roger have decided to improve on this design. They're adding a couple electromagnetic clutches and using an S9G reactor to power the system. They expect to get number-to-number times in the millisecond range, so it wont take more than about 12 hours to run all the possible combos.
Thank you for showing the interior of the lock. I used one of these locks on a gun safe I made for my father, and he accidentally broke the dial off, rendering the lock unopenable. Now that I have seen the internal mechanism, I know how I can defeat it and open the safe.
I spoke with a safe expert. He said auto dialers are dangerous because the more you use a dial, the more risk you can destroy the mechanism, especially with old dials. He said to never use these with old dials, and probably not with new dials if you want to protect the gear. He said even a person trying an old dial a few times can destroy it.
Not that hard to predict when its gonna open looking from the inside, i knew almost exactly when its gonna open by predicting when all three notches will align.
assumedly he already knows the combo and started the machine pretty close to that. Especially since the first number was 63 which would take 20 hours to get to. A cigar and drink is still probably in hand though
I guess since he knows the combination and has a tablet that shows the numbers being tried, he knows roughly how much time it takes to get to correct combo
It's not only a dial safe opener but a doubles as a durability tester. Any safe that endures that and still functions normally would be an industry gold standard of quality.
Newer S&G digital spin locks defeat this attack in a couple of ways. When you start turning the dial in a direction, it picks the starting position at random. Also, consistent spinning of the dial, i.e. not the "pulsed" spinning you get by hand will be rejected.
The newer S&G mechanicals make you push in at 0(zero) and then continue to open also... 1/2 price of the Electronic locks.... and minimal failure rate compared to the Electronics also.
They use the electronic locks on newer SCIF containers (cabinets and room doors). Easier for security to update codes on a regular basis without having a locksmith come in and change 50 container codes.
@@ejm57301 electronic locks are far better in almost every conceivable way, that's why high security ones can be expensive. Combo locks are good for a safe you'll be entering once a week or less. Anything you have to enter a couple times a week or more, there's no better method than electronic.
@@littlejackalo5326 true, but as an EE I know from experience all electronics fail eventually. Not that mechanical locks can’t too but I have more faith in them than a digital lock. Also push buttons can wear leaving a pattern that constrains the number of guesses one would have to make. That said I realize too many wrong guesses results in a delay on most locks…
I imagine it would depend upon the safe being cracked. I've worked with a number of large bank safes in my time that it might struggle with at such speeds due to their poor lubrication and service history. The machine likely has a way to tweak the speed though.
@@Thermalions These NEMA 23 units has a pretty decent amount of torque. They're often used for high end 3D printers. The good thing about the feed back system they're running is that you can set it to "talk back" mode. The program tells it to go to a certain position, and it flags the program when it reaches that position. So you pretty much don't need to account for resistance/slow movement in an application like this.
I want one. Not just the machine, but the safe door to go with it. This thing is so hypnotic to watch. Now I also need to find a video showing the "normal" opening of the lock from the inside to understand it.
LPL, at my work we sometimes install security doors with, I'm fairly certain, S&G 8550MP locks. I recently noticed they had replaced the metal discs in the lock with white plastic ones. When using the robotic safe-cracker, have you noticed any heat build up in the discs? Seeing the change to plastic discs made me wonder if melting or softening the discs would/could be a viable attack. Maybe something as easy as some kind of rubber disc on an angle grinder would spin the wheel fast enough to melt/deform the discs inside the lock. I imagine S&G has put enough r&d into it so that they feel it isn't a problem, but the thought struck me that it might be an option.
What do the discs actually do? Something as simple as butane jet lighter might be enough to melt the disc. Could it be that they are teflon discs? Teflon melts really high and is really slick, sometimes even used as a lubricant.
Can someone make a 10 hour video loop of the robotic safe-cracker outside and inside? it is a really good ASMR almost puts me to sleep when I should be working
I think that you can eliminate a lot of numbers by adding some torque/resistance measurement and spinning the dial. Testing all numbers where there is a sudden change in resistance should point you to the spot where there is a gate so decreased friction on the latch and then trying those in different combinations until it opens.
"Now, normally this security guard would pose quite the problem, but due to an oversight they do not wear any gas mask or other gas attack protection if there is no high security alert. So I'm simply going to toss in a canister of the high grade sleep gas, if you want a canister on your own I sell these on my website."
Until I saw this video, it just never occurred to me that you could check a combination in such a way as to preserve the first two digits while running through all the possible iterations of the third digit.
We have one of these in our shop. Only used it about 3 times in 5 years. Only really need it for Tl-30's that are hard to drill etc. The lock in this video can be manipulated by taking readings easily with practice.
Man ngl, I've struggled with sleep deprivation for years, struggling to fall asleep, but I fell asleep listening to the sound of that machine doing its work. It was oddly enough very soothing and stress-reliefing, my breath synced somehow with the tact. 10/10 would recommend! Very interesting to see how it's working. I would however like to see more of the inner workings of a safe lock, would that be possible? Or is it stupid-simple like, I'm looking at it right now?
Should be split screen showing both the front and back simultaneously. Easy video to make after having set everything up once and people would love it. It's great to watch the back tumbler (or whatever it's called) turning ever so slowly.
Would love to see more vids about the internals of these types of locks. I dont quite get how going back and forth like it does solves the combo. Many of these types of locks I've seen require you to go a multiple rotations past a number.
Note how it does a full sweep of the last number, then increments the 2nd to last number, and sweeps the last again...it's checking all possibilities just not redialing the first 1:2 numbers for checking all of the 3rd.
Back and forth is a trick to exhaustively try the third number after the first two numbers have been dialed. This is due to the mechanical nature of the lock and the only way to defeat it is to get a lock with a 4 wheel pack.
One way to slow down this process would be to make the dial and inner parts larger, thereby increasing the number of intervals/numbers and thus the number of potential combinations (though I understand some people would not want a lock which requires them to remember multiple 3-digit or 4-digit numbers). Another way would be to manufacture the lock in a way that limits the rotation speed of the dial. It could also be designed to temporarily freeze the dial if it is rotated too quickly and require the user to press a button to release it. That wouldn't prevent a brute-force attack from working, but it would slow it down immensely.
Meanwhile, I recently was able to figure out the combinations to three antique safes for a friend in a couple of hours, but that's only because the doors were open and I could access the lock. Two safes I had figured out in about 10 minutes each, but the third took about 1.5 hours, not only thanks to a stiff and worn out (worn numbers/increments) dial, but the mechanism was deeply recessed AND it turned out to be a 4-number starting to the right instead of the left. Although it was frustrating, it was also exciting (sort of) in that I'd only read about 4-number safes but had never encountered one in the wild before.
Unless you're Parker cracking the Bank of England vault, in a very respectable 2h25. Then he cracks their new high-tech vault in seconds with a hairpin 😁
This should also be optimized by executing the least-travel moves first, or at least have the option. I could see it potentially cutting down on time but could also shift a longer travel result further down the brutelist
Great video. Thanks for the inside view - I couldn't comprehend how it was bouncing back and forth between two numbers, not three - 10, 21, 10, 22, 10, 23 sort of pattern - but the inside view clears that up nicely. Could you make a video showing how Hollywood safecracking with a stethoscope works (or doesn't) in the real world?
Running an auto dialer that fast will usually cause the drop in bar to skip over dropping into the drive cam. They also cause quite a bit of wear to the lock and having to run it more than once adds wear and takes more time.
Let it be known that LockPickingLawyer will win ASMR video of the year award.🏆 The cadence of his voice along with the metronome quality of the safe cracking machine make for a perfect video. Thank you!
[1337] new world (LPL's world) record length video! Over 11 mins!! As always, the LPL doesn't disappoint unlike many products showcased on his channel.
why are people acting like it's such a big thing that his video is over 11 mins long lmao. yes, the product here is different than him picking a lock, the video length will be different.
![[1337] My New & Improved Robotic Safe-Cracker…](/img/logo.svg){kind=logo}
