MOD-05

Swarm

WatchingLogistics

Coordinated drone delivery: held as a watching brief, not a build.

A research file, kept current on purpose, not in active development. Moduloa’s active build is Manufacturing.

What this file is

Swarm is where Moduloa watches drone logistics. The idea it started from still holds: delivery is a coordination problem, and fleets need scheduling, queues, and capacity systems more than they need aircraft. What changed is who is building it.

The honest history

Work on adapting early drone platforms for delivery started here in 2019. It did not materialize. That sentence stays on the page: the method that holds this site together documents what didn’t happen next to what did.

The market got built by others

Drone delivery is no longer a gap waiting for a builder. DJI dominates the hardware layer at a scale no small player can approach, and Zipline has passed a million commercial deliveries running autonomous delivery networks as daily infrastructure. This is an industry with giants in it, and pretending otherwise would be exactly the hype this brand exists to avoid.

The position

Watching, not building. Research accumulates here as the market moves, so that if an opening appears at the systems layer, the file is current and ready. Until then, no drone gets built here.

What we monitor

Three things could reopen the door

01 · RegulationBeyond-visual-line-of-sight approvals becoming routine instead of exceptional.

02 · The coordination layerMixed fleets from many operators will need open queues and standardized intake more than another aircraft, the same idea PrintDrop is testing on 3D printers.

03 · Thin rural routesNorway's thin rural routes, closest to the ground Trace already works.

The regulatory state

The rule that would make BVLOS routine is still a proposal

Watching the regulation line above returns this: on both sides of the Atlantic, flight beyond visual line of sight is still permitted case by case rather than by rule, in the US through a waiver or an exemption assessed per applicant, in the EU through an operational authorisation from each national authority.

1,320 lbMax gross takeoff weight proposed for part 108, and waivable

~3,100Comments filed: over half touched right-of-way

89Of 496 BVLOS authorisations surveyed, called economically viable (concentrated in 2 states)

01 · Part 108, United StatesFAA and TSA published their BVLOS proposal on 7 August 2025 (docket FAA-2025-1908, RIN 2120-AL82, official text at 90 FR 38212; the FAA’s own copy linked here is the pre-publication version and says so on every page) to move BVLOS from what the FAA calls “enablement though exemption” to enablement by rule, for aircraft up to 1,320 pounds maximum gross takeoff weight unless otherwise authorised. The Unified Agenda puts it at final rule stage with a July 2026 target, against a statutory deadline of 16 January 2026 under section 930 of Public Law 118-63, now passed, with nothing in force.

02 · What the reopened comment window is actually aboutThe FAA reopened comment on a deliberately narrow list, electronic conspicuity devices, having flagged two provisions as the reason: § 108.195(a)(2), giving part 108 operators presumptive right-of-way over manned aircraft outside Class B and C airspace and away from traffic broadcasting position through ADS-B Out or an approved alternate device, and § 108.180(b), requiring detect-and-avoid of non-cooperative aircraft inside Class B and C. Over half of the roughly 3,100 comments on the original proposal touched right-of-way; others said detect-and-avoid “would require prohibitively expensive technology that has not been adequately proven for safety value.” The notice states that comments outside its listed questions are out of scope. That window closed 11 February 2026.

03 · The two routes that exist meanwhileUntil part 108 exists, US BVLOS runs either on a waiver to § 107.31 or, for larger operators, on a 49 U.S.C. 44807 exemption followed by a part 135 or 137 operating certificate, the route Zipline and UPS Flight Forward fly on, the latter under Exemption No. 18339D. On the waiver path the FAA promises only to “do our best to review and approve or disapprove … within 90 days” (FAA), and its evaluation guidance shows why that is no queue to build on: applicants must show how the pilot continuously knows the aircraft’s altitude, attitude and movement and detects and avoids all other aircraft; sensor-based detection may need certified equipment; and control links on unlicensed FCC part 15 spectrum are treated as insufficient. Neither route is per-flight; both are per-applicant.

04 · Europe’s three categoriesRegulation (EU) 2019/947, applicable since 31 December 2020 across the EU plus Norway and Liechtenstein, sorts operations into ‘open’, ‘specific’ and ‘certified’. Open is capped at 120 m; certified aircraft “will always need to be certified (i.e. have a type certificate and a certificate of airworthiness)”. Delivery lives in specific: an operational authorisation from the national authority covering the flight or flights, unless the operator declares against a European Standard Scenario or holds a Light UAS operator certificate with the relevant privileges, in which case no authorisation is needed at all.

05 · U-space exists on paperThe U-space package, Implementing Regulations (EU) 2021/664, 2021/665 and 2021/666, applicable from 26 January 2023, is already the coordination layer this file argues for. Article 3 makes four services mandatory inside a U-space airspace: network identification, geo-awareness, UAS flight authorisation, and traffic information (Easy Access Rules). But it opens “Where Member States designate U-space airspace”. Designation is a choice. An EASA survey reported alongside its U-space light proposal found 496 of 2,486 active operational authorisations across 11 member states were BVLOS, only 89 of those reported as economically viable, and those concentrated in two member states: Germany and Norway. Trade reporting from June 2026 counts “only one limited U-space area in the entire European Union”, a tally we could not verify against any official register.

What would have to change

Designation is the binding constraint, whatever software exists. With essentially no designated U-space airspace (trade reporting counts one limited area, a figure we could not verify), there is no flight authorisation service at any scale to queue against, and no legal home for the four mandatory services. EASA concedes the adoption problem itself: its Notice of Proposed Amendment 2026-103, in public consultation with comments due 12 July 2026 and described in the trade press, would let member states begin with a subset of the four “where the density of UAS operations does not require strategic deconfliction in space and time.” It is a proposal in consultation, not an adopted instrument, which is the same distinction this section’s headline turns on. In the US the dependency is part 108 finalising, with the third-party UTM approval procedures section 932 of the same statute directs.

The test, stated as a test

Nothing above is a plan and nothing here is being built. It names a condition instead: an open coordination layer is worth attempting when two or more operators inside one designated airspace must share a queue neither owns, and the rule permitting it is in force rather than in a docket. Neither half holds today.

Documentation, decisions, and build logs accumulate here. Everything is public.