This topic contains 3 replies, has 1 voice, and was last updated by 
josh September 3, 2021 at 7:01 pm.
-
Posts
-
joshI imagine a tech like this:
Local elements detect things that look like finger ridges & pressure patterns – their location + 2D + Time is communicated to an integration layer which says yea or nay. Basically. Can it be cheap & good enough to do the job? I bet.
-
joshElaborating – The user has a name, etc. so there can be multiple users. The device ships with a built in sense of what other people look like. There’s a protocol for learning “fred” & multiple trials of [start press,press stop]. Fred works out with the screen when it’s gotten good enough at learning Fred vs. everyone else. There would need to be some instructions & warnings to prevent typical sorts of day to day confounds based on positions or sweat, or mood. Seems like you could make it accurate enough in practic & it’s not that obvious how to hack it at the mechanical level. The wireless level is std wireless security.
-
-
joshAdditional considerations:
Point 1 – Value of adding unique, mobile identities for non-GT members that can be tracked for continuity – biometrics would help. If they later join GT then it’s important to know if it is the same entity.
Point 2 – Using biometrics or other methods, some spooks will try to game the system. In light of that, consider the potential value of mechanism that has hidden levels of escalation which need to be triggered by an third party key. Penetration testing & hacking of the lower levels can’t discover the extra level mechanisms until they are invoked by the third party key. That could give a way to provide extra checking in suspicious or high risk situations. The extra level might reference stored info from earlier.
You must be logged in to reply to this topic.