 for taking your time to being here with me tonight. And also welcome to the people on the stream. This talk is about the Miwa Anker lock. I have one of these locks in my collection. I'm a bit of a lock collector. I used to be chairman of the tool as well. And of course I'm a hacker, like most of you. And I like to figure out how stuff works. And I thought a few years ago already, how does this specific lock work? Because it's quite special in that it uses magnets. So I set up on a trip on figuring out about this particular lock. And this talk, the next 50 minutes, I will tell you about what I've tried to do with the lock where I failed, but not many people will tell you nowadays when we all have Instagram. But I will tell you all the failures and all the successes as well. I probably don't have enough time to have questions in between. If you have any questions, maybe we have some time at the end, or otherwise you can find me outside after the talk. So I have to back route in the tool. And I also have to say I'm writing a book together with a few friends about lock sport. So this cylinder, the Anker 3800 or MIWA 3800 that originated, was invented in Japan as MIWA. It is sold in the Netherlands as Anker. This is a view from the inside. And you see that this lock has four pins, which you will find in many regular locks that you will find on the door. But also here above the pins, we have sliders with magnets in them. And that makes this a special kind of cylinder. Now here you see the key. And on the bottom of the key, there are these indents, the bidding of the key. And you can see that they match the four pins that are inside the cylinder. So and on the other side of the key, we see eight black squares. Those are four sets of two magnets. And those are somerium cobalt magnets. And they can also be fakes. So some of these black blobs are magnets, and some of them are fakes. And these magnets make tiny little sliders in the cylinder move, and they need to move to the correct spot to open. So here again, you see a blown-up version where you can clearly see that the, what did I do? The wrong button, where you can see that the sliders go from left to right within the plug. And they are operated by the magnets that you can see here in the key. And on the bottom, these pins are operated. Now the pins, I'm not really going into lock picking. You can go to other talks or to visit the tool village here at the campsite. But for lock picking or for the pins, they need to be aligned at the shear line. So the length of the red pins need to match the cuts and the key. And when they are all perfectly aligned, the plug is free to rotate. Well, that's not the interesting part. The interesting part is those magnets. So we have here the four sliders that can slide. In this orientation, they slide up and down. And so you see there's four sets of two magnets in the key that one set per slider. Now I then did some photoshopping to create this wonderful slide. So I think this is my Photoshop achievement unlocked. Although it's a bit crappy, but I'm not a designer. I'm a hacker. On the left, you see the closed lock. And you can see that the pin is not at the shear line. And also you see this yellow slider. It protrudes into the edge of the cylinder. So that's why it won't open. And there is a little spring on top. And this green and red thing, that's the magnet. And on the right, you see it with the correct key inserted. So the key pushes the pin to the correct depth. And also the magnets in the key will interact with the magnets in the slider, moving it away to the right, freeing up the plug. So that's basically how it works. And here you see that in this... So I said there are pairs of magnets. But in this case, there's only a magnet on one side of the key. And that is actually quite normal. Here you see a key with a piece of magnetic paper on top of it. And you can clearly make out that there is magnets in there. And I can even also tell that the bottom two, they are either facing north up or south up. And the upper two, they are facing left or right. So there are different orientations of magnets, and that is what creates the key, the code in this key. And of course, you need at least one magnet per slider to operate the slider. If you would use an incorrect key, then maybe the bidding is incorrect, which means that some pins are sticking out, or you have the incorrect magnets, which means that the sliders are sticking out. So they all need to be correct for everything to align within the plug and the plug to rotate. Right, so this is the theory, how this lock works. So I hope that is all clear. Now let's look at defeating this lock. So what about picking? Can we pick this cylinder? Well, the four pins that are in there, they are just four pins. And you could lock pick them with regular picking techniques. And actually, there's only four of them, whereas most common locks have five. So that should be easier. But then there's also these magnets. Now these magnets can also be manipulated if you take a piece of metal and glue a small magnet on the end of it. You can insert it in the lock and wiggle it and manipulate these sliders. And the sliders will move inside of the cylinder. And you can also use audio magnification to make it easier to figure out what is happening and then open this lock. This is not easy. When I started my research, which was already quite a few years ago, I had not seen anybody open this cylinder before. But later on, I found, well, here's at least four videos on YouTube where you can figure out, well, where you can see this lock actually being picked by hand. But there's only a few videos, so that already tells you that this is very hard lock to lock pick because of those magnets. And the oldest videos are about two years old that I found. So what I did was I took an old cylinder and I made a test cylinder out of it. So in one side, I only kept the pins. And the other side, I only kept the magnetic sliders to work with this. And my first thought was, can we duplicate the key? Because you cannot just duplicate such a key. You need to call to the factory to get it duplicated. You need to have a certificate. But if you have a key, can you make a copy? And of course, you need to read out the magnets. So we need some kind of device to measure the magnetic field. Now, I'm quite an old hacker. So I like to do electronics, real old stuff, electronics. I have a talk about the gigatron tomorrow night. But that's a different story. So I made this. I found this on the internet. And I made a magnetometer. So here's a device that you can use to measure magnetic fields. And as you can see here, I can read out the magnets. And you will see that some magnets are orientated left to right. And some of them are up down. Oh, that's not true. In this particular key, they're all left to right oriented. And you can also see that for each pair, there's only one that is an actual magnet. The other one is fake. OK, reading the magnets, done. My electronic badge is in the pocket. But now I need to create a duplicate key with these magnets. Now, the blank keys cannot be obtained. So I cannot start off with a blank key and put in magnets and make it fit. The magnets are made by a factory in, I believe it is in France or Spain or somewhere. And the only company in the Netherlands that has these is the anchor factory itself. And they won't just give you a key. Maybe if you're very good in social engineering. But anyway, I thought I'd make a magnet setup key. So I take an existing key, try to take out all the existing magnets. So I have holes in which I can put my own magnets. So I use my Dremel. And I also took away the bidding. And my idea was this key is used to set the magnets. And then I can use this slit in between to pick the pins. Because, yeah, I cannot form an existing key. I cannot make another bidding of another key. So I thought, well, let's make a key with the magnets and do picking. Well, that was also an epic fail because there's so little room inside this cylinder that I just could not put my pick in the lock and not disturb pins that I did not want to move. So that was a fail. But nevertheless, I proceeded. I bought some 2 by 2 by 2 millimeter magnets. And I tried to make the setup key anyway. Because I wanted to do, yeah, I still was trying to learn more stuff. But this just didn't work. I'm not very good at metal work, it appears. And it was too steep a learning curve for me. I didn't have the right equipment. So the setup key, a thing was going nowhere. It failed. So I thought, how else? Well, let's forget about these magnets for a while. Let's suppose that we can create a key with the correct magnets. Let's suppose I find somebody who can do the metal stuff and create this empty key for me. If I have it, what do I do with the bidding? How do I copy the bidding? Well, I also do have some equipment to make copies of keys by using molding. So you take some putty, you put in the key, you get a mold, you pour in mold metal, and you have a copy. Here you see a mold I made of an existing key. So now my idea was, if I have a setup key with the correct magnets, I need to put in this bidding that you see here. And this is how I did it. So the setup key, I combine with the mold of the key I want to copy. I pour in metal, and this is the result. Does it work? Yeah, yeah, achievement unlocked. Of course, this is the cylinder that only has the pins and not the sliders, because I didn't succeed in making the setup key. But this proves that if I have a setup key, I could copy the bidding. So to duplicate a key, we could make a skeleton key from the existing key, but that's quite frustrating. Also, even if I could get this empty key with the holes to put the magnets in, the magnets that I bought are two by two by two millimeters, and the key is only 2.1 millimeters thick. If you have only 0.1 millimeter left, it becomes really horrible to work with. And this was a state of affairs at LOKON 2019, just before COVID. And I thought, well, the only solution I see to advance is to do 3D printing. But I didn't have a 3D printer. And yeah, I left the project. I abandoned it. But then came COVID. And I had time. And I also had a little bit of money left to buy myself a present, which is this Creality Halo printer. This is a resin printer, because I thought because of the really tight tolerances, I needed a very printer that could make really minute details. I bought this printer, and I thought, and I was still somehow stupid of me in the mode of I need to get that setup key. So what I did was I made a setup key with the slit to do the picking, which I already knew didn't work anyway. But this was fun because I needed to learn about open-escat and cat design. So again, achievement unlocked. I was able to create a 3D model of this key. And this has the slit to allow for picking. And this is actually maybe the second or third thing I printed on this printer. So I was really chuffed. It did need a little bit of filing to make it really fit, because the measurements were a little bit off. But as you can see, the magnets did fit. It didn't break. And best of all, it does work on my cylinder that has just the magnets. So the proof of concept is I can make this key. Still a tiny problem of the slit where I cannot put in my picking tool without doing anything. But first, I fixed my model to make it fit better. So this is a somewhat later key. And this fits right away after printing. But I still had that problem. In the meantime, I had learned from those YouTube videos, which were made after I picked this project up again in COVID. I found out that the pins are first to bind. There are many locks with different rows of pins of different elements. You need to pick one row first. Then the lock will rotate slightly. And then you need to pick another one, the second row of pins. Now, in this particular lock, you need to pick the pins first. Then it will rotate a little bit. And then you need to put the sliders in the correct position. So I can actually, by applying tension in the middle of the lock and using a picking tool, I can pick the four pins. And when the four pins have been picked, I can use the setup key that I made by copying the magnets from the original key. I can insert it and keep the tensioner in place. And that would open the lock. And I tried this and I picked the lock. And actually, it was much harder than I had anticipated. But achievement unlocked. I did open it. But this is not really something that a non-professional lock picker would easily do. Also, the pins in this lock are anti-picking pins, which makes it more difficult to pick them. Yeah. But as I said, oh. So the attack factor that we have up till now is if I have access to a key, even temporary, I can read out the magnets. I can create a setup key. I can pick the pins and open it. But of course, why am I not just 3D printing the bidding? I was really stuck on my original train of thoughts of continuing what I was doing. But of course, you can just print the bidding. I mean, that's no biggie. So I updated my model. And you can now create bidding in there as well. And you can just type in another code. And as you can see, it will change. There's another bidding. So you can just type in the code of the bidding that you want to have. And you create a key. So with that, here's such a key that has been printed. And it does work on both sides. So now we can actually create a copy. Of course, you also need to know what the bidding is. But there's only four pins. And the depth of these pins, there's only four possibilities. So that's not much. In many logs, you will have seven or maybe nine possibilities. And then you really have to measure quite well to see which depth it is. But in the anchor, it's only four. And if you've been doing this for a bit of time, you can do this with a naked eye. If you give me an anchor key, I can see with a naked eye what the depth are. So I can put that in my open SCAT, print a key, and have a duplicate. So that's interesting. The problem is a little bit that the keys are very brittle. I once dropped a key from this height and it broke in two. And also, you have to be careful inside of the lock. So I also did have a key break inside of the lock. And it gave me very hard time getting that key out again. So that's also why if you do these kinds of experiments, you do them on your own cylinders that you don't use in a door or somewhere where you rely upon them. So the keys are brittle. So what to do about that? I thought about this. And I thought, well, if we make the magnets smaller, there's more surrounding material making the key more sturdy. So I did find some 1 by 1 by 1 millimeter magnets. Now they are so incredibly small that if you put them in tweezers and you move them around, they will already flip. It's, I gave up. And then I talked to a colleague of mine at work about this hobby project. And he said, well, why not use this magnets? And he actually already started on printing on his printer a key. And he does not have a resin printer, but a filament printer. And as you can see here on the bottom, this is a key where the magnets are cylindrical. This means that some cylinders, some magnets need to go flat and some magnets need to be going deep inside of the key. But of course, if you print a key on a per use basis, you can print it any way you like. So you just print it so it fits the correct magnets. And what's even better, you can leave out the holes where you don't need a magnet, where we have these fake magnets, which will also strengthen the key more. And I was worried about the filament printer not being able to make the key with such tolerance that it would work, but actually it does work flawlessly. So that's about key copying. But another kind of attack is where you want to gain entry to a, or gain entry, open a lock where you do not have the original key. So this is not about key copying, but about opening the lock, decoding, picking, whatever. Now here are again some wonderful photoshopping. And you see that the key on the right has a magnet with north facing up. And this matches for this slider, a magnet where north is facing right. And it always has to be the same. If you have this type of magnet in the key, you know what needs to be in the slider to make it move in the correct direction. And for each of these magnets, there's five possibilities, north up, north down, north to the left, north to the right, or no magnet at all. So here you see the arrows represent the direction the sliders need to move. So two of them move right and two of them move left. And that's the same for every cylinder. And here you see the magnets that I've read out. And I then know if I see this key, I know already that the lock must contain these sliders. Because if this key is inserted in this specific lock, I know that it's the only way that the sliders will be moved to the correct position. So knowing the magnets in the key, with that I will learn what the magnets are in the cylinder. But it's also the other way around, if I know what is in the cylinder, I then know what is in the key. So I wanted to read out the magnets in the cylinder. So I used my wonderful magnetometer, tried to read out the magnets, but yeah, it doesn't fit. The sensor is too big. As I said, this lock is as a weird shape and there's really not a lot of space. So this was a fail. And I looked for smaller ones, smaller whole sensors. And I found this in a SOT 23 package, which is rather small. I bought a few of them. And then I needed to put this on something to insert into the cylinder. I then tried the key cat to design a little PCB, just a flat piece of PCB that goes inside of the cylinder. And I bought some 0.4 millimeter flexible PCB. And I was told by somebody you can, so that's a key cat achievement. And somebody told me, if you print your key cat design with a laser printer on inkjet paper and then use an iron, you can transfer it to the PCB and etch it. Well, for me, this failed horribly. I don't know what I did wrong. Maybe some of you know, come see me after the talk. But I reverted to really old school PCB making, which I've already done years ago. It's just using a Sharpie and drawing on it and making it. So that was achieved, but it still failed because it didn't fit. Here you see on the bottom left, you see that the sensor is on top of the PCB. And together with the PCB, it was still too high. So the solution was to put the hole sensor at the end of the PCB and glue it on. And for that, I needed to extend the wires. So I used 0.1 millimeter wires and soldered them on. And this was the most hard thing of the whole project. It was really frustrating, but it did work in the end. So it was on, it was working. I used it with my magnetometer and it failed because it uses different voltage levels than the other hole sensor. So I needed to fix my electronics, but I thought, well, let's learn some more stuff. Let's do something with an Arduino. I mean, it's totally overkill for reading out the hole sensor, but it's fun anyway. So here you can see the device in action. So I can insert it. I have to wiggle it a little bit in case the magnet is going left and right. But you see I can actually read out the magnets that are in the cylinder. And what's interesting is that this lock has quite a few magnets, but we'll get to that later. And since I was doing this anyway, I thought, well, let's make it more beautiful. So I ordered this LCD touchscreen and I also programmed in that you can, that it will translate from the magnets in the key to the cylinder and vice versa. It has now an analog mode and a digital mode to show what the magnetization is. There's an SD card in there. So you can save the current readout to the SD card. And here you see me reading out a specific cylinder. And you see it will translate automatically. If the cylinder in this position has this kind of magnet, then the key must have this other magnet in that place. And now if I read out the original key, I should get what is depicted on the left. Works great. So I can now read out the magnets in a cylinder. So the attack factor is if I have temporary access to a cylinder, but not the key, I can read out the magnets and I can determine what magnets need to be in the key. But I don't have the bitings. So still I can make a setup key, but I still need to do the lock picking for the four pins for one time opening with a lot of effort. Or I could make all possibilities because it's only four pins with four possibilities. If I spend a few days and make 256 keys, I could try them out and one of them will fit. So do we have an alternative instead of picking or creating 256 keys? So one of the ways that you can also open a lock one time is using bumping. So I won't explain everything with bumping, but the idea very shortly is that you apply a lot of force to the pins that the force is moved over to the other pins on top. They will separate for a short moment of time and in that time you can open the lock. Can we do bumping? Now that's quite hard as well because for bumping you need a bump key. A bump key is a key where as you can see here in the movie the key is cut to the deepest most position everywhere. But for this I need an existing key. And I also need to have the correct magnets in there. So I could print a... So yeah, I cannot get a blank key from the factory because they won't give it to anybody. But I can print a key. Well, the resin printed key is much too brittle. So if I smash it, it will break. But fortunately the Pat G key is strong enough. So this is Pat G on the filament printer. So I have the correct magnets that I've read out. I use a bump hammer. And of course I show you the video where this works in one hit. Normally it would take maybe five, six, seven, eight hits to open it. It does require a bit of practice to be able to do this. But it works and it did work repeatedly. So the key was not... It was still operational for another bump after I did this bump. So that's nice. It gave me the possibility to do a one-time opening if I don't have the key. There's another way of finding the bidding which is impressioning. That's also a very interesting topic that you could also talk about for 50 minutes. Instead, maybe you can watch this video by Jols Weijers. He did a great job on attacking master key systems. How to do that with impressioning. With impressioning, you start with a key that has all the pins in the top, maximum level. You wiggle it in the lock and the pins will make tiny marks on the key. So you need to have a key that has rather soft material like brass, not steel. And you get tiny marks. You need a magnifying glass to see them. And where you see a mark, you need to file away a little bit of material until it fits. Because if you file it to the correct depth, it will no longer mark. And here you see a lock. So here's a dimple lock key that I impression. And you see it looks a bit weird because I filed straight over the key, but it doesn't matter as long as the place where the pin drops goes to the correct spot, that is fine. Now, can we do this for the anchor? Well, not with our 3D printed keys. I mean, that material is not suited for doing impressioning. You need to apply quite a lot of force and these pins need to mark the key. So we need a brass setup key. And we don't have those. We have a brass setup key, a brass impressioning key. Well, I was talking to my colleague, Hido, about this project, and he said, oh, but I have a CNC machine. Maybe I can do something for you. So actually he did an aluminum key for me as well. So from that one, he took some pictures. So here you see that he's creating from a piece of aluminum or aluminum, I should say, probably here, made a key with the spaces to put the magnets in. And you can make this a bump key or an impressioning key. And here's the impressioning key. So this was made out of brass. And on the bottom left, you can see that under the right lighting, you can see the marks that the pins make on the key. And there you file away a little bit and you try again. So this is not something that you can do in a few minutes. It takes several steps and quite a bit of time. And yeah, did I achieve it? Yeah, I achieved impressioning a key to make it fit. But I feel I also kind of failed because this was very hard. I mean, I've done impressioning in the past with regular locks and the locks you saw before. That is way more easy than with this particular lock. I'm not really sure what is the reason for that. Maybe it's also the material used, but I found it very hard. And it's not really a real life scenario, I would say. But still, we now have an attack factor when you have access to the cylinder, but you don't have a key. We can read out the magnets. And we could create a set of key with the deepest cuts to do a one-time bump. So you have opened it, but you don't know what the combination is. Or if you're very experienced, you might be able to do impressioning, which will take quite a bit of time. But then you end up with knowing the correct bidding. So you have an actual key for a cylinder that you have not seen the key for before. Now, let's talk about something else, which is master keying. Because the anchor lock is a high-security lock. It was invented in the 1980s, I think I already mentioned. It was patented in the early 80s. The patent is no longer valid by long time. But they are used in high-security environments like hospitals and such. And in those environments, they use master keying a lot. And a very short introduction to master keying. Master keying is where you have cylinders that each have their own key. So you have a key that works on your door, in your office. But you also have a key that works on all the doors on your floor, or all the doors in your building, or all the doors on the side. But let's keep it simple, and I'll show you an example with just two locks, with two individual keys, and one third key that opens both of them. How does that work in a traditional system? This is a traditional pin tumbler lock, master keying. So on the left is Alice's lock, and on the right is Andy's lock. And here you see Alice's key. Alice's key will open Alice's lock. Why? Because all the pins are here straight at the shear line. So that works. But it doesn't work on Andy's lock, because as you can see, some pins are still blocking. So Alice's key only works on Alice's lock. Andy has a different key, and here you can see that Andy's key will work on Andy's lock. But it does not work on Alice's lock. So far, so good. And now comes the special key, the master key, or grandmaster key, or submaster key, or whatever you call it. And this key will open, as you can see, both locks. And you've also seen that the way they do this is by cutting up the pins once more, so you have two possible double-set that pin to the correct position. Well, there's two correct positions. And what's also interesting to note is that the master key that you see here on top in this case, in this example, differs in two spots. And the master key always has more material on it than the individual user keys. And if you're a real hacker, you will probably figure this out. But the reason is that if you have the user key, you must not be able to create a master key by just filing away a little bit of stuff. So you can create an individual key by the master key by filing, but that's not really an attack factor. Now, if you translate that to Anker, then you will see that also there we have master keying in the pins. So the pin in red is actually, it's now lying flat, but it is a really thin spacer. And on the middle right, you see the three different spacers that exist with a depth of 1, 2, or 3. And you also see the pins with depth 1, 2, 3, 4. So with these, you can make all kinds of combinations. So for this, these pins in this cylinder is part of a master key system. And there is two bitings that will match to open the lock. So my naive thought when I started this was to think, well, if I have an existing key, so let's get back to the scenario where we do have a key for a cylinder. And we know it's master keyed. And I have this key. I know it's a user key, so I don't have much access. Then, well, this must be the correct position also for the master key. Because otherwise, I could file away stuff to get to the master key, which is against our rule. And this pin, well, the master key might have one of three other positions there. And this one, and we can try them out. So we create, we print three keys. And we can just alter one of the pin stack positions. So we keep three positions the same. And the other one, we can vary. And we make three keys, and we test them all. And maybe none of them work. That gives us information. And for the third key, there's two more possibilities. And we print them, and we try them out. And maybe only one works. And for the last one, we print three. And also, maybe one works. And naively speaking, we would say, well, it must be the highest positions for everyone. That must be the master key. And we can print that. And hopefully, that will open all of the locks. It is actually not true. Oh, yeah, we can copy these. But you can make these key keys, try them out, and get information about what works. What was interesting is that we do not have just the pins. We also have the magnets. And that makes this lock much more interesting. And in the magnets, we can also do master keying. Because the magnets, here you see the bottom of the slider. You see there's place for two magnets. Also, in the key, we have place for two magnets. And this we can use for master keying. So we can have a key with a magnet on the left side and another key with a magnet on the right side. And the slider has magnets on both sides. And that's really cool in this lock. Here's a key. This is a key that's part of a master key system. This is a user key. If I have this key, I can read out the magnets. And I already know, using my little device here, it will tell me what magnets are in the cylinder. So, but when I, let's go again. So I know at least these are in the cylinder. And, but what is in the master key? Well, what I can do is I can also read out the magnets in the cylinder in my door. And maybe I will find some extra magnets that are in the cylinder, but they do not match anything in my key. Now, so that gives information. So first of all, I know that the master key, the master key operates on my door. So it must have the magnets to operate, it must have magnets to operate each slider. Now there are two sliders, the middle ones, that only have one magnet. So I know that the corresponding magnet must be in the master key. So that is something I know for sure. Now there are also magnets that are in the master key that I did not find corresponding magnets in my key. But there are in the master, in the, oh, I said master key, right? The magnets I find in the cylinder that are not matched on the key, but they must be in the cylinder for a reason. I mean, you could frustrate hackers and put in magnets that do not do anything, but I'm pretty sure that's not the case. So these magnets are there, they are operated by some other key other than my key. So that must be the master key. So the master key must have magnets that operate on the blue magnets in the cylinder. What to do with these sliders that have two magnets? So I still have the north up here and I have the south up there. Should we put the corresponding magnets on the master key or not? Well, I can give you an example that shows you that you should not. So let's suppose that, so this is the key we just saw and this is my cylinder. But in this master key system, we also have a different cylinder, cylinder B, with its own key, key B. And if you look closely, you will see that key A does not open cylinder B and key B does not open cylinder A because here is a south matching the north south and here is a north matching the south north. So they can all be mixed. If the master key would have a north south there or a south north, it would not open one of the keys. So if we go to cylinder B, read out the magnets in the key, we find out that on the bottom left, there's a north. We know that in the master key, there should be no magnet. That's something we learn. And we also learn by reading out the second cylinder that it has a magnet that we had not seen before. There's a, on the second slider, there's a north on the right. And again, this slider only has one magnet. So the master key must operate that one magnet. So in the key, there must be the corresponding magnet south north. And for the one on the top left in the master key, we are not sure yet. We do not have enough information. It could be that there is no magnet in the master key or it could be that there is one, but if there is one, it must be a north south one to operate on these two cylinders. So this is a process of looking at keys, measuring the magnets, measuring the magnets in the cylinder, going to a different cylinder, measuring the cylinder to a different cylinder. And each time you see something that you hadn't seen before, you get more information to paint a complete picture of what the master key will look like as far as the magnets go. And the last one, the last magnet, well, we didn't see any magnets, neither in the key or in the cylinder, so we just don't know yet. But you could make a key with these magnets and see if it works as a master key. And if it doesn't work, you need to read out more cylinders. So getting a master key, well, what I said before about filing away, that was a bit naive. Also in the anchor system, because you have both pins and magnets, that means that if you have a key that could be filed to the master key bidding, that's not an issue if the magnets are different. Because if you are an end user and you file away, well, you do not have the capabilities to exchange the magnets. At least I don't, I don't have the CNC stuff. So if you take your key to another lock and you wonder if it's the master key and it doesn't work, well, that's a bit of an issue. You can read out the magnets in the cylinder to figure out what the correct magnets are, but you cannot figure out what the correct bidding is. So what do you need to do? What you can do is to check all other possible biddings in the original lock, so you have the first lock with your own key. And for each pin, you try the other three possibilities and that will give you information about all the possibilities that are present in the master key for the bidding. And then you can make a whole lot of keys, but not 256, but at most 80, depending on the master key system, and try them out and try which fits in the second lock. So that's interesting. The more cylinders you see and the more cylinders you measure, the better you are able to get the master key, or as I call it, gold mode, the key to all, but it's surprisingly hard. Just, and the reason is this combination of the pins and the magnets. So I'm actually quite impressed with this system that is already 40 years old and it still has such a resistance against finding out the master key. It certainly is a lot harder than with regular pin terminal locks only. So let's summarize. I may even have time for a few questions I see, which is great. It's normalizing here for the people who are more visually. If you start, if you have a key, if you don't have a key, now let's start with not having a key. You don't have a key, then you can decode at least the magnets in the cylinder so you know which magnets are in the key. Then for the pins, well, either you can pick, but that's very hard. That's why it's in light blue. I don't think that's really a realistic way of doing it. You could bump. I think bumping is quite realistic, but then you have a one-time opening. If you are able to do impressioning, which is also very, very hard, you could get a working key. So getting a working key without having one at start is very hard, but opening once is for a very determined hacker, doable. If you already have a key, an end user key, then you can decode the cylinder and the key. It gives you more information. You can duplicate the key. And if it is master keyed, you can try with creating a lot of other, well, with the 12 keys, find the other bitings, and then you can go to other cylinders, read them out. And maybe you need to print quite a lot of keys, but if you are persistent, you will be able to, in the end, find the master key, but that is still quite hard. Now there's one thing I didn't really discuss, which is that if you have opened the cylinder once, what you could also do is open the cylinder and look at what's inside. So you don't have to make all the setup keys, but you just open it up, look at the pins and the depth, and you can maybe reassemble it or throw it away. But this is also quite hard. If you have a cylinder like this, well, at least I'm not able to disassemble this, but I'm able to disassemble it, but I'm not able to then reassemble it. If it would be a half cylinder that is doable, still quite hard. So that's actually where I am with the investigation, it took me quite a lot of time. I did have a bit of a wish list, but I didn't have time enough to implement this before MCH, so I wanted to build an Arduino shield where you have 16 hole sensors, so you can put on a key and get a reading immediately for the complete key. But obviously it doesn't really add to what I've been saying, but it's just a nice project to do another PCB and some electronic stuff. And what also would be really nice is to make a device that would read out all the magnets from a cylinder in one go. And I did later find out that there are devices that magnetometers that are suitable for that. But yeah, this is harder to solder and I didn't really pursue this yet. So the conclusion with the advent of 3D printers, copying keys has become much more easy. For regular standard keys, this becomes super easy. And yeah, if you show your key to somebody and it's a regular key, somebody can just make a picture of it and make a copy and 3D print it. With these, it's still a bit more complicated with these keys because if you have a picture, you could see the bidding, but you can't see the magnets. But of course, if you have the bidding, you can go to the cylinder, it should, the key fits in and read out the magnets, and then you have everything as well. So you should still always keep your keys in your pocket or at least non-visible to others. But the attacks are also in this law quite difficult because there is such a constrained space. Find the grand master key. I thought it was, it would be quite easy, but in the end, it is quite complicated, much more than with conventional locks. And please do not test stuff on cylinders that are not yours or that you do not use. And in the end, I would really like to spend a minute on Hito, who's helped me out quite a lot, a colleague of mine, but also Ankerslot. So in the process of responsible disclosure, I talked to Ankerslot a while ago saying that I was going to do a presentation and I wanted to hear their thoughts about it and they were really super about it. So the lead developer invited me to the factory and yesterday. He showed me around, we talked about this, you knew about these things and he was really a great sports about it. So I would like to, if he's watching, I would like to thank him very much. And in the end, I would like to thank you for spending your time to being here with me tonight. Thank you very much. Thank you. We've probably got time for maybe one or two questions, if anyone's got any. Oh, I've got one quickly. Could you, could you've used electromagnets instead of putting magnets in? I have contemplated electromagnets, but it was too much of a hassle to find, to make or find electromagnets and do it. And yeah, it would be very interesting because you could make a sort of setup key and maybe flip through all the possibilities and do brute force attack or just click and you have the key. But it would be just too hard for this project to actually make. I'm already very happy with the result I have so far. I spent a lot of time on it and I learned a bunch of stuff about a lot of things and that's where it ends for me now. Oh, super, thanks. Does anyone else have any other questions? Oh, we've got one here. Since you've talked to Anker as an ethical hacker, is there a recommendation you have for them to improve their lock so it's even harder to pick? Well, they're already, I would say pick proof. I mean, for a company that buys a lock that wants to deter burglars, this is secure enough, this is fine. And they're still selling these cylinders and they sell quite well, I believe. But of course, they're also working on new ideas. So I talked to the developer, they also put magnets in their standard line. They have an infinity line that also has a magnet and they're also making some, they are actually making some improvements to this specific cylinder. So yeah, there's still a bit of development going on there. But otherwise, yeah, I think this, and you should also not change too much on the design to prevent introducing other flaws. So maybe they will come up with a completely new design. Okay, great. Well, thank you very much indeed for explaining that to us. It's fascinating listening to you.