You seem to be arguing for overengineering without considering the costs. Should the failsafes also have failsafes, should the backups have their own backups? And what if the additional security measures add additional risks? Take #4, "human patrols". Unless implemented very carefully, greater access for security personnel could well add more risk than benefit.
I'm not even sure that one should argue that the vault failed from the point of view of the owners. It was designed well enough for them to do their daily business and such that they were able to obtain insurance on the contents. A vault doesn't have to be impregnable, only visibly difficult to steal from.
> A vault doesn't have to be impregnable, only visibly difficult to steal from.
I'm pretty sure a vault in this situation has to be really impenetrable and definitely not just visibly difficult to steal from.
What you're proposing is the exact situation they had. The result didn't look good.
To obtain better security, the motivation of the participants needs to be analyzed. Perhaps this is a problem best solved through obscurity, with two competing security teams installing two totally separate alarms systems, each standing to lose if their alarm is defeated. Or perhaps skip the alarms, and rely on threats of deadly force.
But although any new insurer will properly demand the technological improvements illustrated here, it would be foolish to believe that technology alone will keep out a dedicated team of insiders with full access to the system. Real world security is not equivalent to public key encryption where security can be established by 'proof'.
1 - The magnetic plate sensors should NOT be accessible while the vault is closed. There is no reason why this should have been mounted outside the vault instead of inside.
2 - All sensors in the vault were tied to a single circuit . Shorting just two wires was enough to take every single sensor in the room offline. That's ridiculous.
3 - You shouldn't have a simple "break circuit if alarm tripped" design at all. You have a full-blown sensor package mounted on a wall in a secure vault, is it too much to ask that the sensor respond to pings? This is a fairly minimal cost (+$100 per sensor, maybe?) that would also make it impossible to short the sensor circuit or otherwise take a sensor offline without somebody realizing.
thiis the #1 thing that would prevent a lot of crime. people dont do it because it makes logistics harder. people like routines.
If nothing else, it would help detect a failing component.
How do you lock up the 'impossible to duplicate key' - by definition you would need a lower-level security key to secure it?
Spoilers below if you haven't read the article:
1. The key shouldn't be kept in the closet outside the safe
2. Opening the door should force the magnetic sensor to be triped, the door must be in the way of the sensor. Additionally, why was the sensor on the outside of the door?
3. The combination lock should have a viewing angle that requires the user to block any cameras.
4. You need human patrols.
5. Sensors should be independent; the heat/motion sensor shouldn't both need to be tripped.
6. They needed sensors that could not be easily seen or detected by someone in the safe (hidden cameras, sensors embedded in another object).
7. The needed multiple sensors instead of one each.
8. The sensors should adjust the current flowing through them so that the wires can't be bridged, and the wires shouldn't be exposed.
9. There needs to be a random factor inserted such as random patrols, a sensor that moves in a random way, the combination is changed randomly.
10. They should have a time control on the safe that only allows it to be opened during certain times. There could be an override, but it'd be it should be difficult and rarely used (hard to copy).
Just a few off the top of my head, more cameras would also be good -- what else did they do wrong?