In 2019 we changed Vizzbee's UTM architecture from one central control plane to region-partitioned UAS service suppliers. Each supplier was responsible for a geographic area and exchanged the information needed for handoffs at the boundary.

The standards chronology matters here. We were working from UTM and USS concepts already circulating through regulators and industry groups. ASTM F3411-19 formalized Remote ID at the end of 2019. The interoperability specification now known as ASTM F3548-21 was published later; our architecture subsequently aligned with several of the roles and interfaces it formalized. The 2019 design did not begin from a finished F3548 standard.

What the central service would have owned

A central service could keep the complete operating picture in one place: aircraft positions, intended routes, restrictions, and conflict-resolution decisions. It would also become the party that every operator had to trust with that information.

That trust assumption became difficult once we considered the participants that might share the same airspace: commercial operators, public agencies, and equipment from vendors we did not control. The architecture needed an answer for who could read shared state, who could change it, and what happened when participants were governed by different rules.

We did not have a convincing operator for the center. That governance problem mattered more to us than a generic scaling argument.

What federation changed

Partitioning by region reduced the amount of state any one supplier needed to hold. It also created the possibility of limiting a failure or compromise to a smaller operating area. That was a design objective, not a property we had formally proved.

Federation introduced work that the centralized design avoided. Boundary handoffs became part of the safety case. Suppliers needed compatible interpretations of shared messages. Conformance testing mattered because independently operated systems had to behave consistently at the interface.

We chose federation because the ownership model of the centralized design was not credible for the airspace we expected to serve. The later standards gave that decision a clearer interoperability framework; they were not the origin of the 2019 design.