Do you have any details / references on the specification?
Do you have any details / references on the specification?
I believe this is a draft [0] (it could be final too at this point). From there, here’s the specific ID randomization:
> UTM (UUID): A UTM-provided unique ID traceable to the Registration ID that can act like a “session id” to protect exposure of operationally sensitive information.
[0] https://www.reginfo.gov/public/do/eoDownloadDocument?pubId=&...
From your reference:
UAS ID Type 1. Serial Number, 2. Registration ID, or 3. UTM U
(page 10)
That is: there is a single identifier, constant over time, that is assigned to each drone aircraft.
The UUID standard appears to be RFC 4122:
https://www.rfc-editor.org/rfc/rfc4122
And again: so long as that's transmitted in the clear, ANY party may accumulate a database of UUIDs and locations over time, which would tend to identify the party(ies) associated with a specific drone. That might not be precisely the operator, but if drone ID "8daf5e654951a033175fb71e0a2ecc03"[1] keeps showing up at Your Fine City's Police Department ... or sewage department, or electric utility ... you might begin to draw some inferences regarding its likely owner/operator.
It might be easier to do so if the identifier had, say, a specific entity element to it (say, as with LDAP or X.500 entity attributes), but the lack of such an element to the identifier itself is at best a small hinderance to owner/operator identification.
________________________________
Notes:
1. md5sum of the randomly-selected string "random drone id"
Which is required to defeat any level of time-based observational collection and correlation by third parties. As is commonly the place with present aviation tracking systems.
https://www.eff.org/pages/automated-license-plate-readers-al...
Again: once an identifier is available, correlating it with other information is trivial.
With nothing more than two postal codes (ZIP codes in the US), representing home and work, it's possible to individually identify about 90% of the population. That information is available via, say, geocoded location using mobile phone or tracking cookie data.[1]
In the case of device or vehicle tracking such as with drones, it's highly likely that a given device would be used within a particular jurisdiction or activity region. If you can pinpoint specific locations and times within those (e.g., police activity around an address + drone activity, activity following specific infrastructure such as power lines or gas distribution, training locations, etc., etc.) you can probably come up with a strong idea of who operates or owns the equipment.
That is more-or-less the process that was used to identify CIA-operated aircraft used in renditions for "enhanced interrogation" (torture) by amateur plane watchers. See: https://en.wikipedia.org/wiki/Rendition_aircraft https://web.archive.org/web/20130616175922/http://www.common...
________________________________
Notes:
1. The home+work ZIP anecdote is the one I recall. I'm not finding that specifically though this paper mentions four spatio-temporal locations sufficing for 95% of the population: Yves-Alexandre de Montjoye, César A. Hidalgo, Michel Verleysen & Vincent D. Blondel, "Unique in the Crowd: The privacy bounds of human mobility", Scientific Reports volume 3, Article number: 1376 (2013) https://www.nature.com/articles/srep01376
I'm sure there's some kind of hypothetical downside, but not one that me or anyone I know have encountered at least.