Project Caribou — Phase 1 Review & Phase 2 Proposal
Phase 1 ended with July. All build and electronics goals are met. The hover milestone moves into Phase 2 — deliberately, under a proper legal test setup. This post reviews Phase 1 against its defined goals and proposes an adjusted Phase 2 scope. After this discussion, a Snapshot vote formalizes the transition.
Full Phase 1 detail lives in the monthly reports, most recently the July 2026 progress report. This post only summarizes results against the goals.
1. Phase 1 review — goals vs. results
The original proposal defined Phase 1 (“Proof of Concept & Electronics Integration”, months 1–4) with these elements:
Goal: Build the test frame — steel core, detachable aluminum motor beams.
Done. We assembled PT1 completely: six X15 powertrains with 18S Tattu batteries, full avionics. First weigh-in: 182 kg dry.
Goal: Develop the electronic system and integrate it into the test frame.
Done, deeper than planned. We designed, manufactured, and installed all three Caribou PCBs. We wrote the CBC firmware. All six packs report full telemetry to the flight controller. The CAN architecture is validated on the aircraft.
Goal: Power-on tests and a general test of the electronic system.
Done. We ran full-system powered ground tests and a complete motor test on July 31. All six drivetrains matched within ~1% RPM. Zero CAN errors. The logs produced a verified first-flight parameter set.
Goal: Design studies for scalable frames, alongside the test drone.
Done and ongoing. Three frame concepts were planned as discussion material: Nils’s commissioned study (delivered), KBM’s design suggestion, and my own concept (in progress). Review runs in the design threads.
Milestone: stable tethered hover flight.
Open. The aircraft, parameters, and flight checklist are ready. What remains is the test setup itself: the site (approved in the last week of July) needs preparation, a 1-ton tether anchor, and the SORA 2.5 application — a fast-track process that still takes a few weeks. We chose this path over a quick unofficial hop. It gives Caribou a legal basis for 12–24 months of test flights, and it means we can show exactly how a 182 kg open-source aircraft is tested responsibly. We therefore propose the phase vote now, with the hover as Phase 2’s opening milestone.
2. What Phase 1 taught us
- Weight. PT1 is 182 kg dry, ~40 kg over target. At the motors’ continuous rating, every kilogram of structure costs one kilogram of payload. Weight reduction is the highest-value design lever for any follow-on frame.
- Transport. The rigid arms overhang the trailer. Folding arms are an operational requirement, not a nice-to-have. We will fit the sourced folding connectors to PT1 and gather real feedback before designing them into a next frame.
- The regulatory path is concrete. With the secured test site and the SORA 2.5 application, flight testing becomes a 12–24 month campaign and also secures fast testing for future Prototypes.
3. Proposed Phase 2 — adjusted scope
The original Phase 2 (“Scalable Design Overhaul”) planned to transform the test frame into a deployable prototype, build a second prototype in the USA, and integrate payload systems.
The intent stays. The sequencing changes. PT1 is a flying asset — we should extract maximum learning from it first. The second build becomes an improved PT1, not a clean-sheet redesign. Flight experience then informs the next mayor prototype build instead of preceding it. Four steps:
1 — PT1 flight-test campaign.
- Fly the tethered hover (the carried-over Phase 1 milestone).
- Then advance in capability stages, not flight hours:
- Tuning: stepwise controller tuning until PT1 holds attitude and position reliably. The tuning documentation is itself a deliverable.
- Small automated missions in the tethered setup: short waypoint missions, automated takeoff/landing, position-hold tasks.
- Endurance and lifting tests: sustained hover runs and lifting tests against the motor ratings.
- File the SORA 2.5 application and transition from tethered to official free test flights.
- Install the folding motor-arm connectors for road transport.
- Deliverables: flight logs, tuning documentation, endurance/lift results — published to the repo.
2 — PT1.5: adjusted design, build, and flight (USA).
- Adjust the PT1 design in a small, deliberate step — no clean-sheet redesign: proven core architecture and electronics, folding arms from the start, the weight reductions that need limited re-engineering, lessons from PT1 assembly.
- Two standard payload bays: compartments dimensioned to accept DJI Agras T40 tank modules( or similar) — liquid spray or seeding tanks drop straight in. Proven, mass-produced payload hardware instead of custom tank development, ready for the first real use cases (agricultural dispensing).
- Build and fly this PT1.5 with Thomas in Texas. This fulfills the original Phase 2 goal of a US build for manufacturing feedback and flight experience, at low engineering risk.
3 — Upgrade PT1 to the PT1.5 standard.
- After PT1’s initial flights are done, retrofit it with the PT1.5 adjustments so both aircraft converge on one configuration.
- This keeps PT1 flying as the test aircraft, proves the adjustments work as a retrofit and leaves the project with two matching aircraft instead of two divergent prototypes.
4 — Documentation that enables a community build.
- Bring the PT1.5 documentation (BOM, build guide, configuration, test procedures) to the level where a community member can build the aircraft.
The frame-concept comparison (Nils’s study, KBM’s suggestion, my own concept) continues as a design discussion — mainly through August, then it will naturally quiet down. Its results feed the future full redesign. We deliberately attach no Phase 2 milestone to it; the full redesign build belongs to a later phase, informed by the flight data and payload experience this phase produces.
Phase 2 milestones:
- PT1 tuned and demonstrating small automated missions in the tethered setup, with endurance and lifting results documented
- PT1.5 designed, built, and flying in the USA with the dual payload bays
- PT1 upgraded to the PT1.5 standard
- Documentation published that enables a community build of the aircraft.
4. Timeline and budget
- Phase 2 window: September 2026 – February 2027 (months 5–10, unchanged).
- Budget cap: unchanged at $30,000/month (labor $25,000 / hardware $5,000), no rollover, per the original proposal and AIP-007.
- The prototype construction in the U.S. will likely be carried out in collaboration with JPL (Javelina Propeller Laboratory). Arrow will not have to bear the main costs for this prototype, but may provide some individual components.
5. Governance
This thread is the public phase review required by the project proposal. After discussion and the customary sentiment poll, a Snapshot vote referencing AIP-007 formalizes the Phase 1 → Phase 2 transition with the scope above. Bring comments and objections here.
Proceed to Snapshot
Hold up, needs clarification
Reject Phase 2 proposal
Abstain
Sources: original project proposal, July 2026 progress report, project repository.