31

Every time I visit a clinic, office or train station I notice how easy it is to figure out the PIN required to unlock the staff-only door by just watching an employee entering the restricted area.

I usually don't have to do anything, not even trying not to get caught, simply it's pretty easy to know the PIN from a decent distance because the keypad design in typical and everybody knows the location of the digits and which one corresponds to which square.

However I went to the bank and the keypad there had a different design which made it difficult to guess the PIN. What is the down side of this design? why don't we see it used in safes?

enter image description here

  • 15
    How, specifically, does this pinpad operate? Is it just the same as a standard numeric pad with the buttons in a circle rather than a grid? Or is there something more clever going on? – Polynomial Apr 21 '16 at 11:03
  • 29
    So instead of static number pads, these are dynamic displays? That would make sense because you would have to spot the numbers they are pushing and not just the location. Kind of like a phone lock screen with random number patterns, such that the finger prints don't give away the digits. – Potaito Apr 21 '16 at 11:15
  • 12
    Reminds me of how the bank PIN works in Runescape. Every time you press a digit the locations of each number changes, and having your mouse over a key hides the number. – Ajedi32 Apr 21 '16 at 16:20
  • 11
    Also, unlike with the static design, all buttons will be used the same, preventing you to find digits in the PIN by checking for wear, smudges or fingerprints. – Edheldil Apr 21 '16 at 16:23
  • 1
    And if you enter the PIN based on the location of the previous button permutation does security get alerted?! – MrWhite Apr 22 '16 at 15:09
33

I can think of several possible downsides to this sort of circular dynamic display:

  1. The dynamic display slows down the user entering the access code. Because the numbers will not be located in the same place every time the user needs to stop and read the keys before pushing the appropriate button. This may take only a second or two longer than a traditional keypad but additional time could allow a second pair of eyes to better observe the number layout as well as the keys punched.
  2. The dynamic display requires the user read the keypad. With static number pads, everyone knows the layout so it is easier for a person to enter their code without looking at the key numbers. This means that a person can more closely stand to the keypad as to block the view of any "peeping Toms". With a dynamic keypad the user must more closely read the keys which means that they will naturally stand farther away from the console, allowing more space for wandering eyes to observe.
  3. The circular design spaces the keys farther apart than a square layout. You mentioned that you can intuitively tell which buttons are being pushed based on the known static keypad layout but assume you didn't know where 0-9 was located on a square keypad. One of the nice things about a square keypad is that the numbers are bunched together so an individual can somewhat block the view of the keypad with a hand while pushing buttons. With a circular design, your hand would be more splayed out making it much more obvious which buttons are being pushed than a square pad. This weakness is clearly mitigated by the dynamic numbering but I can't help wonder if a square dynamic pad would have been better for concealing the buttons being pushed.
  4. It can be argued whether this is necessarily a weakness but I do notice that this keypad clearly has an electrical dependency. Mechanical locks have their own strengths and weaknesses but it is important to note that electrical locks can be vulnerable to power outages and surges as well certain types of electrical tampering. In that same vein... these types of electrical doors are typically hooked up to a network which allows for computerized vulnerabilities as well.

As mentioned in other posts, there are a whole lot of advantages to this design as well, though, so I suppose with this keypad you are trading fingerprint smudge vulnerabilities for "looking over the shoulder" vulnerabilities which could easily by limited with proper training.

  • 1
    I'm not aware that mechanical keypads were a thing that is normally used. – Tomáš Zato Apr 22 '16 at 10:33
  • 1
    @TomášZato They are more common in older 70s-era buildings in the US. They typically are built directly into the doorknob. You can find a bunch on Google Images. A lot of the ones pictured are actually electrical but the older mechanical versions I've seen used push-buttons that I believe turned rotaries within the mechanism to unlock the door (just as a grooved key does). – DanK Apr 22 '16 at 12:12
  • 2
    Additionally to your vulnerable to power outages point: those LEDs can fail and arguably will before a normal key would wear out. Extra maintenance, etc. – Jared Smith Apr 22 '16 at 14:05
  • 1
    I work in a building that uses a square grid random layout number pad to unlock some doors. It has an additional mitigation to hinder shoulder surfing. Restrictive viewing angles; at arms length the area that the number's are readable is only about 1.5-2 heads wide/tall. In some locations this is farther mitigated by angling the pad up slightly because more than a yard or two behind you the viewing cone is above head height even for a basketball player. From the outside I believe ours are designed to fail locked; for what they protect a delay to call a locksmith is acceptable tradeoff. – Dan Neely Apr 25 '16 at 5:06
57

There is a Security User Experience (SUX) downside, which you might consider to be minor.

As someone who is more kinestheticly inclined, I don't memorize things like phone numbers or PINs: I memorize patterns. If I was forced to use this keypad, I would have to use a compensating method to remember the actual digits (like writing it down).

While not everyone is strongly kinesthetic, it is a factor in how people learn, which means that others might need to write down the passcode, too. This, of course, defeats the purpose of having a dynamic keypad.

As I say, it might be a minor point, but it is a downside that impacts security.

  • Is there any research as to the propensity of users towards kinesthetic, vs other approaches? (what are the other approaches?) – goodguys_activate Apr 21 '16 at 15:54
  • 3
    @LamonteCristo Learning Psychologists have been trying to keep track of that, and there is a lot of research. "VAK" or "VARK" should result in some research articles. – schroeder Apr 21 '16 at 15:58
  • 1
    @CortAmmon Not everyone will be so security conscious, though, and "writing down passwords" is, imho, one of the worst possible outcomes of a security policy. Some people will even stick them to the device they use them against (e.g., a monitor). I think it's unlikely a sticky note to a lock in a clearly public area would last, but if a user thought the area was secure, I could see it happening. Bottom line being that I wouldn't consider this a minor issue. – jpmc26 Apr 21 '16 at 21:02
  • 2
    @jpmc26 Depends on what kind of security you are going for. Is the password there to deter your co-workers / housemates or attacks over a network? In the latter case you would be undoubtedly more secure to write down a random 20ish character password and stick it to the monitor. Targetted and random attacks have different (and to some degree opposed) prevention strategies. – Jared Smith Apr 22 '16 at 14:01
  • 1
    Hold on, I'm kinesthetically inclined and I can form all sorts of patterns on that thing. It would only trip me up if I had to enter the same password on keypad with a different design. Infact this is likely to keep my from reusing a favorite pattern from a typical pad. Which is a good thing – candied_orange Apr 23 '16 at 3:33
18

There pretty much isn't a downside, aside from cost and design complexity. From what you've shown it looks like the numbers randomly cycle across the buttons, so there's no set pattern to watch. These types of keypads are quite common in prisons, where the pads are quite close to cells or are in communal areas (e.g. the dining hall or laundry areas) and sneaky prisoners will watch the guards day after day to guess the pin pattern.

I'm not sure if the circular design helps, but I guess it's slightly harder to identify which button was pressed. I can't say I've got anything to base this on, though.

As for why they aren't commonly used in safes? Cost and simplicity is probably the biggest factor, but also the fact that safes usually aren't repeatedly opened in front of potential threat actors (e.g. in a public space) but are rather in a closed off area such as your house, or in a back room of a bank.

  • 2
    "There ... isn't a downside" is contradicted by the other answers. But the rest of what you say makes sense. – Toby Speight Apr 22 '16 at 10:47
13

Other answers have addressed most points that occur to me, but there's one more I can mention: it has accessibility problems for users with visual impairments. Standard keypads have a locating nipple on the central button (normally the "5"), and users can enter the code using offsets from that position.

With a dynamic arrangement of input positions, then even if the centre can be successfully located, then the user must determine which button is to be pressed (I'm guessing there's no Braille on the buttons).

Many places have laws on accessibility that would make such devices problematic for public use (e.g. payment card terminals); similarly, employment laws could also force companies to make alternative arrangements for any of their staff who are unable to use them.

5

Quick guess would be that before displays were cheap, a quick and effective way of changing the buttons pattern was to have the digits printed on a disc and physically rotating it to change the digit shown in each window. Doing that on a square keyboard would be much more of a hassle!

  • I saw a conventional grid layout with a 7-segment number on each button (LED or VFD) on the entrence to a building in the 80's. I think this would be practical as long as electronic keypads have been practical. – JDługosz Apr 21 '16 at 18:40
  • 1
    @JDługosz This idea is interesting because it could be implemented completely mechanically. Given a source of mechanical energy - such as a wound spring or a gravity-driven weight drive - the "scrambling" of the digits could be done without electrical power. Add an interlock to the door itself to raise the weight or wind the spring and this would be self powered. A standard dial lock is simpler, and spinning the disc each time would slow the process down, but still a fun mechanical solution. – Corey Apr 22 '16 at 0:45
  • @Corey Doesn't even need a separate source of energy. Use pushing the button to cycle the labels. Biggest issue would be randomizing purely in mechanical hardware... – Someone Somewhere Apr 22 '16 at 12:22
  • rather then rotate the displayed numbers how about rotating the clear cover that is over the buttons. Some people insist on keeping the same pin for so long that it starts to wear marks into the pad. – candied_orange Apr 23 '16 at 3:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.