2026‐06 - Arrow-air/project-caribou GitHub Wiki

June 2, 2026

Caribou Call — Recording · HearHear transcript

Attendees: far1no, kbmollysuh, errrks.eth, alperenag, zeynepb5793

• CBC PCB / Enclosure Model Handoff

  • Julius said the CBC PCB should be ready for enclosure work and believes the relevant model is in Fusion, but KBM saw an older C90/CBC version and missing top-edge clearance in the current Fusion assembly.
  • Julius said the old PCB model is still usable for some enclosure-design work because the missing top clearance does not change the basic enclosure shape, but KBM may upload/update the latest PCB part himself.
  • KBM estimates the enclosure work will take roughly one full week after the model is updated.

• Enclosure Design Constraints

  • Waterproofing is not required for the first prototype; Julius said dust protection is enough because Caribou will not be flown in rain.
  • KBM flagged sealing, heatsinking, and maintenance access as likely design challenges.
  • The PCB component side currently faces the airframe, making maintenance harder if the enclosure must be opened.
  • KBM suggested flipping the PCB so components face the battery side and can be accessed through a lid; Julius said this could be changed later, but the first prototype can continue with the current orientation.
  • Direction for PT1: continue current PCB orientation, leave extra space around the heatsink area in case fan/passive cooling changes are needed, and avoid over-optimizing waterproof access for this revision.

• Connector / Sealing Details

  • KBM asked how the large output connector seals to the enclosure.
  • Julius explained the connector has a sealing/gasket face intended to seal against the enclosure, using screw holes for compression.
  • The enclosure may need upper/lower split geometry or a lid split around the connector area if the team wants stronger sealing later.
  • For PT1, the connector can remain somewhat maintenance-unfriendly if it is still workable.

• CBC PCB Review

  • Erick has reviewed about half of the CBC PCB so far, starting with the power-regulator section.
  • The regulator section looks good and appears to follow datasheet example layouts; nothing urgent has stood out yet.
  • Erick will focus next on the kill-switch / latching / trigger circuitry and then continue through the rest of the PCBs.
  • Julius explained the kill-switch latching mechanism is intended to keep the main MOSFETs engaged even if the ESP32 reboots, and to prevent the ESP32 from independently disabling the MOSFETs. Disabling should only happen through the 12 V kill-switch signal from the main PCB or the battery button.
  • Battery workflow: plug in all batteries, enable them one by one, regulators start, ESP32 closes the main MOSFETs after a few seconds, precharge is disabled, then the battery can only be disabled with the battery button or key switch.

• Main PCB Power Architecture

  • Julius is mostly working on the main PCB and hopes to have an initial version ready this week.
  • The main PCB will combine 12 V sources from the batteries into three redundant 12 V sources, each combined from two batteries.
  • From those three 12 V sources, the main PCB will create three separate 5 V circuits that can be assigned to different devices or sensor sets.
  • The flight controller can use two of the 5 V circuits for redundancy, leaving an optional third 5 V rail for other sensors/peripherals.

• May Report / Contributor Updates

  • Julius still needs to create the May Caribou report and plans to do it this week.
  • KBM can send screenshots/monthly records of his work to make the report clearer.
  • Julius said KBM can share PCB updates with Vector for GitHub import, or create a branch/zip for Julius to review before a PR.
  • KBM asked whether Vector can export a full PCB STEP file for Fusion; Julius said yes, that workflow should work.

• Additional Clearance Concern

  • KBM flagged a possible main-PCB output-cable clearance issue where a metal beam may block a large connector/cable path.
  • Height clearance is roughly 8–9 cm and may still be workable, but the team should check with the updated PCB inserted.
  • KBM suggested that the next prototype may need flipped PCB sides and possibly shorter main-PCB dimensions.

• Open Items

  • Julius/KBM: confirm Fusion contains the latest CBC/C90 PCB model or upload the updated part.
  • KBM: begin/continue CBC enclosure design once the model is updated; rough estimate one week.
  • KBM: design PT1 enclosure for dust protection rather than full waterproofing, while reserving space for future gasket/sealing features where practical.
  • KBM: leave extra heatsink/fan clearance in the enclosure design.
  • Julius/KBM: keep current PCB orientation for PT1, but consider flipping component access toward the battery side for the next prototype.
  • Erick: continue CBC PCB review, focusing next on kill-switch/latching/trigger circuitry.
  • Erick: continue review of other Caribou PCBs after CBC review.
  • Julius: continue main PCB work and target an initial version this week.
  • Julius: update README/documentation items that are now stale after Vector’s earlier work.
  • Julius: create the May Caribou progress report this week.
  • KBM: send screenshots/monthly records of his Caribou work for the May report.
  • KBM/Vector/Julius: use Vector to import KBM PCB updates to GitHub or create a branch/zip for Julius to review.
  • KBM/Vector: export PCB STEP files for Fusion if needed.
  • KBM/Julius: check the possible main-PCB output-cable / metal-beam clearance issue with the updated PCB inserted.

June 9, 2026

Caribou Chat — Recording · HearHear transcript

Attendees: errrks.eth, far1no, thomasg, kbmollysuh

• Main PCB / Battery Connector PCB

  • Julius is still working on the Caribou main PCB. The schematics are finished and he is now aligning/placing everything on the PCB.
  • He expects it will not take much more time and wants it ready in the next few days.
  • The main PCB will be populated with many waterproof vertical connectors, similar in concept to a modified Quiver PCB that sits centrally and connects outside sensors/peripherals.
  • Besides the connectors, the board mainly contains voltage regulators and battery/ESC connector interfaces.
  • Julius found small improvements needed on the battery connector PCB routing.
  • He found at least one capacitor with the wrong voltage rating and wants to change two MOSFETs used around the isolated DC-DC converter to smaller/more appropriate switching parts.
  • The current design reused MOSFETs similar to the larger solid-state switch MOSFETs, but those are not ideal for the isolated DC-DC converter because they are relatively old-school / slower switching.
  • Julius plans to document the specific part issues and changes.

• Peripherals / Enclosures / Sensor Mounting

  • Julius has received the flight controller and GPS hardware, and ordered another camera and HM30 unit.
  • The first prototype will reuse several Quiver-style peripherals because that is the fastest path to testing.
  • KBM asked about vibration isolation for the flight controller; Julius said the approach is like Quiver, where damping would need to happen at the big PCB/enclosure level.
  • KBM suggested a future small PCB for the flight controller/sensitive sensors connected flexibly to the main PCB, but Julius said that will not be done in this version.
  • KBM said the CBC enclosure is about 90% done.
  • Julius said the main PCB enclosure still needs to be designed after the battery connector case is finished and parts can be ordered.
  • The frame also needs mounting points for the F9P GPS units and at least the two radar sensors used on Quiver.
  • KBM asked about an aircraft-wide covering/enclosure for dust/splash protection from the ground; Julius said most sensitive components are already in enclosures, so there is no full-aircraft enclosure plan for now.

• Foldable Arms / Payload Integration

  • Thomas asked about foldable motor arms. Julius has not made more progress since identifying RGA-X Hobby as the likely candidate.
  • RGA-X Hobby has a 50 mm horizontal folding mount and a 60 mm downward-folding mount; Julius wants to ask whether they can make a 60 mm horizontal version.
  • Cost is high, roughly $350–$500 per unit.
  • Thomas also asked about cargo/agricultural payload integration.
  • Julius has not designed mounting points yet beyond the early render with two tanks.
  • Thomas noted agricultural payloads such as grain/liquid likely need bottom dispensing and top loading, while cargo may need a study of expected payload types.
  • Julius agreed it would be useful if someone could spend time on payload integration, but his focus is currently on PCB/airframe work.
  • Thomas said Arrow may need to recruit for that kind of work.

• Longshot Bounties / Contributor Sourcing

  • Julius said the Longshot bounties have not attracted much response after more than two weeks, so he reached out to someone on Upwork.
  • He found someone willing to take on the 3D-printed parts and expects those bounties can be marked claimed/in progress.
  • PCB-related bounty work remains open.

• GPS / Heading / Compass Strategy

  • Julius ordered two F9P GPS units; the vehicle can use one like Quiver or use two for GNSS heading by mounting one at the front and one at the back.
  • KBM said a front/back dual-F9P layout gives a long baseline and reliable heading source.
  • Thomas noted compass calibration will be harder on a large aircraft, but dual GPS yaw should reduce compass-interference risk.
  • Direction: mount two F9Ps front/back for heading and also keep the Matek compass, likely disabling the Matek GPS itself.
  • Thomas recommended keeping a usable compass fallback because GPS loss without a good compass could be dangerous.
  • Large Vehicle MagCal plus MagFit from flight data was mentioned as the likely ArduPilot compass workflow.

• CBC Kill-Switch / Latching Circuit Review

  • Erick asked Julius to clarify the CBC kill-switch / latching circuit.
  • Julius explained the 12 V trigger signal path: the main PCB routes 12 V from the battery through a relay path, and when switched it returns 12 V as the control/trigger signal.
  • Erick confirmed the Schmitt trigger input cleans/noise-filters the signal and adds delay, then drives MOSFET control logic.
  • The key clarification was that U10 is normally closed; the trigger signal deactivates/opens the MOSFET switch. Erick had assumed it was normally open, which caused his confusion.
  • Julius also explained that the latch holds state if the ESP32 crashes and that startup reset defines the power-up state so the circuit boots in the off/safe position.
  • Julius and Erick agreed notes should be added/copied into the PCB documentation to make the normally-closed behavior clearer.

• Decisions / Direction

  • Main PCB schematics are effectively finished; Julius is now focused on placement/alignment and wants the PCB ready in the next few days.
  • Battery connector PCB will get small quality/routing and component changes, including capacitor voltage correction and more appropriate MOSFETs for the isolated DC-DC converter.
  • PT1 will keep the flight controller/sensitive sensors integrated with the main PCB/enclosure approach; a separate flexible flight-controller PCB is deferred to a later version if needed.
  • No full-aircraft dust/splash enclosure is planned for now; sensitive components should be protected by their local enclosures.
  • Foldable arm direction remains RGA-X Hobby if they can provide a 60 mm horizontal folding mount, but cost is high.
  • Payload mounting/integration is an open design area and likely needs a dedicated contributor or study.
  • Caribou should use two F9P GPS units front/back for GNSS heading, while keeping the Matek compass as a fallback and likely disabling the Matek GPS.
  • CBC kill-switch documentation should be improved to explicitly call out normally-closed behavior and startup/off-state logic.

• Open Items

  • Julius: finish main PCB placement/alignment and target a ready-to-review version in the next few days.
  • Julius: update battery connector PCB routing/components: capacitor voltage rating and isolated DC-DC MOSFET choices.
  • Julius: document the battery connector PCB changes / parts that caused issues so Erick can track them during review.
  • Erick: continue reviewing the CBC and related PCBs, including the kill-switch/latching circuitry.
  • Julius/Erick: add clearer notes to PCB documentation explaining normally-closed MOSFET switch behavior, 12 V trigger path, latch behavior, and startup reset/off-state logic.
  • KBM: finish the CBC enclosure from ~90% to ready state.
  • Julius/KBM: after the battery connector case is finished, plan/order parts and begin main PCB enclosure work.
  • Julius/KBM: add/plan mounting points for F9P GPS units and at least two Quiver-style radar sensors.
  • Julius: contact RGA-X Hobby about whether they can provide a 60 mm horizontal folding motor-arm mount.
  • Thomas/Julius: consider recruiting or scoping a contributor task for Caribou cargo/agricultural payload mounting and dispensing/loading integration.
  • Julius: mark the Longshot 3D-printed-parts bounty as claimed/in progress with the Upwork contributor, if appropriate.
  • Thomas/Arrow: help amplify remaining open Longshot PCB-related bounties if still desired.
  • Julius: implement two-F9P front/back GNSS heading layout and keep Matek compass fallback while disabling Matek GPS if appropriate.
  • Thomas/Julius: use Large Vehicle MagCal and later MagFit from flight data for compass calibration strategy.

June 16, 2026

Caribou Chat — Recording · HearHear transcript

Attendees: errrks.eth, far1no, kbmollysuh, thomasg

• Main PCB Status

  • Julius reported that the Caribou main PCB is now fully routed with no DRC faults on his side.
  • He still wants to review and clean it up, but the board is effectively routed.
  • The board has many waterproof connectors plus power-supply circuitry and otherwise resembles the Quiver-style central electronics approach.
  • Julius imported the main PCB into the Fusion assembly so enclosure work can start.
  • Julius’s latest routed main PCB is not yet in GitHub because the prior archive/import attempt had file problems.
  • Julius will zip and send the updated files to Vector again so GitHub can be updated; Erick will review after that.

• Main PCB Enclosure / Damping

  • KBM will continue enclosure work using the Fusion assembly.
  • The main PCB is roughly 210 × 220 mm, plus extra space for connectors and adjacent units.
  • Julius said it should still fit on the printer in one piece and does not need to be watertight around every connector.
  • The mounting holes match the Quiver enclosure pattern, so KBM may reuse/extend the Quiver enclosure concept with a different lid or expanded body.
  • The enclosure should mount around the center/crossing beam, likely using the existing four holes and possibly a clamped half-section printed part with rubber in between.
  • Current clearance between PCB and beam is about 3 cm, but Julius said it can be modified.
  • For damping, Julius is thinking about rubber dampers between enclosure and frame plus smaller/softer dampers between PCB and enclosure, similar to Quiver.
  • KBM suggested damping-board or gimbal-style isolation for the PCB and continued to prefer a future standalone/floating flight-controller structure.
  • Julius agreed that can be considered later, but PT1 should use enclosure/frame damping plus PCB damping.

• CBC / Battery Holders / Peripheral Mounts

  • Julius plans to test-print the CBC case soon and test fit when ordered parts arrive next week.
  • Julius ordered stainless-steel battery holders for the first prototype because he was concerned aluminum might be too soft or otherwise unsuitable.
  • KBM noted stainless will be heavier, but Julius said weight is less critical for this prototype and stainless is safer for the first test.
  • After CBC/main PCB enclosure work, next packaging tasks are front/back F9P GPS holders, radar mounts, and camera mounts.
  • KBM said those peripheral mounts should be easier than the CBC enclosure and will collect the relevant 3D models and documents.

• Foldable Beam Connectors

  • Julius found foldable beam connectors on Alibaba and ordered six pieces.
  • They are cheaper than the RJX Hobby options discussed earlier.
  • KBM still noted that this class of hardware can be expensive.

• Camera Mounting

  • Camera mounting should stay lower priority for now.
  • Julius initially wanted the SIYI A8 mounted higher on a tower-like structure for better view over the aircraft.
  • Julius also wanted two fixed cameras so operators can see motor arms or other views in the dashboard.
  • KBM raised vibration/oscillation concerns with a tall camera tower and suggested a GoPro/selfie-stick-style mount or broader aircraft extension bar.
  • Erick and Thomas pushed toward not over-investing yet; Erick suggested a GoPro-style test camera unless a video stream is immediately needed.
  • Thomas said camera mounting is probably not the most important item right now and has spare R1M cameras available if Julius wants free cameras.

• PCB Review Handoff

  • Erick feels mostly good about the CBC PCB review.
  • He still wants to double-check traces and pin connections to avoid past mistakes, but otherwise CBC review is nearly done.
  • Erick opened the main PCB CAD files, but will wait for Julius’s latest routed files to be imported to GitHub before reviewing the main PCB.

• Decisions / Direction

  • Caribou main PCB is fully routed with no DRC faults on Julius’s local copy, pending cleanup and GitHub import/review.
  • Main PCB enclosure work can start from the Fusion assembly; it does not need full watertight connector treatment for PT1.
  • Use a practical damping stack for PT1: enclosure-to-frame damping plus PCB-to-enclosure damping; defer a standalone/floating flight-controller PCB concept to a later version.
  • Stainless battery holders are acceptable for PT1 even if heavier than aluminum.
  • Peripheral mounting priorities after the CBC/main PCB enclosure are front/back F9P holders, radar mounts, and simple camera mounts.
  • Avoid over-optimizing camera tower/video-stream integration right now; GoPro/R1M/simple test-camera mounting is enough unless a specific dashboard/video need appears.
  • Erick’s CBC PCB review is nearly done; next major review is the updated main PCB once Julius gets the routed files to Vector/GitHub.

• Open Items

  • Julius: review/clean up the fully routed Caribou main PCB, then package and resend the updated files to Vector for GitHub import.
  • Vector: import Julius’s updated routed main PCB files to the Caribou repo once received and resolve any file/archive issues.
  • Erick: finish CBC PCB review by double-checking traces, pin connections, and any missing/unconnected items.
  • Erick: review the Caribou main PCB after the updated routed version is in GitHub.
  • Julius: test-print the CBC case and test fit once the ordered parts/battery holders arrive next week.
  • KBM: continue/design the main PCB enclosure using the Fusion assembly and Quiver-compatible mounting-hole pattern where useful.
  • Julius/KBM: decide the PT1 damping layout: enclosure-to-frame rubber and PCB-to-enclosure softer damping.
  • KBM: collect 3D models/documents for front/back F9P mounts, radar mounts, and camera/peripheral mounts.
  • KBM/Julius: design front/back F9P holders and radar mounts after or alongside enclosure work.
  • Julius: test the ordered stainless battery holders when they arrive and confirm fit/stiffness.
  • Julius/KBM/Thomas: keep camera mounting simple for now; consider GoPro/R1M/selfie-stick-style temporary mounts before designing a tall dedicated camera tower.
  • Julius: evaluate the ordered Alibaba foldable beam connectors when they arrive.

• Open Questions / Risks

  • Will the updated routed main PCB archive import cleanly into GitHub/KiCad without the file problems from the previous attempt?
  • Does the PT1 damping stack provide enough vibration isolation for the integrated main-PCB/flight-controller approach?
  • Are the stainless battery holders stiff enough and geometrically correct despite the weight penalty?
  • How much clearance should KBM reserve between the main PCB/enclosure and the center beam once damping and connectors are included?
  • What minimum camera setup is actually needed for first tests versus later dashboard/video-stream goals?

June 23, 2026

Caribou Chat — Recording · HearHear transcript

Attendees: errrks.eth, far1no, kbmollysuh

• CBC Enclosure Fit

  • Julius test-printed one CBC PCB enclosure and said the printed shell, lid fit, gaps, and connector output area look good.
  • He used limited tree supports and has not yet tested the enclosure with physical PCBs or screw inserts.
  • KBM said the design goal was minimum support and was pleased with the first-print result.

• Main PCB / CMAIN Review State

  • Julius told Erick the current GitHub version of the main PCB has only minimal updates and is ready for review.
  • Erick said he will get on the review.
  • Julius has pre-ordered most PCB parts and wants to place the PCB order this week.
  • Expected PCB lead time is roughly 10 days; Julius plans to assemble other hardware while waiting so the vehicle can come together quickly after PCBs arrive.

• Main PCB Enclosure / Service Access

  • KBM flagged a tradeoff in the main PCB enclosure: adding top clearance can move connector cutouts farther from connectors and make locking tabs harder to reach.
  • Current flight-controller-to-lid clearance is roughly 4.3-4.4 mm.
  • Julius thinks that is acceptable for the first print/test and can use a screwdriver if fingers cannot reach connector locks.
  • KBM may lower the lid by about 1.5 mm, but the direction is to avoid complex shaped/split-lid redesign for the first test.
  • The design continues to use the same/current rubber dampers to keep the BOM simple.

• F9P / Frame / CAD Packaging

  • Julius added F9P GPS placement to the frame.
  • The F9P mounts still require drilling holes in the beam; Julius expects careful alignment and already needs drilled holes for battery mounting and other frame hardware.
  • Julius will continue adding hardware to the CAD model and plans to add simple 3D cable-route shapes for high-voltage cabling.
  • Erick offered to model high-voltage cable routing in Inventor if Julius wants help after the CAD model is ready.

• Raspberry Pi Compute Module Direction

  • Julius changed the Raspberry Pi approach on the main PCB to use a Raspberry Pi Compute Module with an adapter board mounted into the same header.
  • This saves side space and changes Ethernet/USB access orientation.
  • Ethernet latch access may be awkward and could require a tool or unplugging the adapter.
  • The Compute Module chip faces up toward the lid, which should make future ventilation or heatsinking easier.
  • Longer term, Julius thinks the Compute Module carrier could be integrated directly into the main PCB using open-source carrier-board designs.
  • KBM still prefers future flexible-PCB/flexible-cable approaches for separating Raspberry Pi, flight controller, and main PCB modules, but Julius noted that this is a different direction from the current integrated-board plan.

• Connectors / HM30 / Antennas

  • KBM suggested using more JST-GH 1.25 mm connectors in future internal designs, including parallel pins if current per pin is not enough.
  • Julius has received waterproof connector samples and wants to test whether they are practical to crimp and insert.
  • KBM designed two pairs of possible telemetry antenna cable exits and a tight HM30 clamp with tape space because HM30 has no screw holes.
  • Julius asked KBM to modify the antenna holes toward a Quiver-like screw-in antenna arrangement, one on each left/right side, without antenna extension cables.
  • KBM warned that large batteries may block or degrade signal and still recommends extended/fastened antennas if needed, but agreed to try the Quiver-style geometry using Quiver design data.
  • Julius will test-slice/print the lower enclosure part after KBM's update to check HM30 fit. The quick slice showed about 7 hours print time. Julius plans PETG-CF; KBM recommended regular PETG but said CF may be okay.

• Decisions / Direction

  • CBC enclosure first print is promising; validate with real PCBs and screw inserts when available.
  • Current main PCB/CMAIN GitHub state is ready for Erick review.
  • PCB ordering is targeted for this week.
  • Keep the current main-PCB enclosure lid clearance acceptable for first print/test unless access proves bad.
  • Use Raspberry Pi Compute Module plus adapter board for this version; consider direct carrier integration later.
  • Try Quiver-style left/right screw-in HM30 antenna holes, while treating signal quality as a risk.

• Open Items

  • Erick: review the current Caribou main PCB / CMAIN GitHub PR.
  • Julius: test CBC enclosure with physical PCBs and screw inserts once parts are available.
  • Julius: place the PCB order this week if remaining review/order checks are acceptable.
  • Julius: continue CAD packaging work, including high-voltage cable route shapes.
  • KBM: adjust the main PCB enclosure/HM30 area for Quiver-style left/right screw-in antenna holes.
  • KBM: preserve enough HM30 access for cables and indicator lights while keeping the clamp tight.
  • Julius: print the lower HM30/main-enclosure part after KBM's update.
  • Julius: test crimping/inserting the waterproof connector samples.

• Open Questions / Risks

  • Will 4.3-4.4 mm flight-controller-to-lid clearance leave enough practical connector access?
  • Will direct left/right HM30 antennas work with large batteries nearby, or will extended antennas be needed?
  • Does the Raspberry Pi Compute Module adapter leave enough Ethernet/USB service access after enclosure install?
  • Are the waterproof connector samples easy enough to crimp and insert repeatably?

June 30, 2026

Caribou Chat — HearHear transcript

Attendees: kbmollysuh, far1no, errrks.eth

• PCB Orders

  • Julius ordered the Caribou PCBs.
  • The order includes 10 assembled battery connector PCBs, 2 assembled main PCBs, and 20 unassembled 90-degree adapter PCBs.
  • Julius estimated the battery connector PCB order at roughly EUR 1,500 for 10 pieces including parts he had already pre-ordered; the bare PCB portion was around EUR 372 before assembly/components.
  • Heavy copper and assembly both contribute to the cost.
  • The manufacturer is expected to finish in a little over a week and ship around the end of next week if the schedule holds.
  • Julius chose assembly for the battery connector PCBs because hand-soldering about 6 needed boards would be time-consuming and annoying.
  • The 90-degree adapter PCBs were ordered in quantity because the price difference between 5 and 20 pieces was small.

• Phase 1 Flight Target

  • Julius said July is the final month of Caribou Phase 1, so the team needs to make a flight.
  • The first test path is conservative: a normal ground test first, then a tethered hover-style flight.
  • Julius described the Phase 1 objective as one hover flight; KBM also remembered hovering, turning, and shifting left/right, but the exact acceptance scope needs confirmation.

• Test Site / Transport

  • Julius has not locked a test site yet.
  • He expects to transport Caribou on an open trailer using a Chevrolet Suburban and wants to keep the route local over nearby country roads.
  • Candidate locations include open space near wind power plants where Julius previously made a PT2 flight.
  • KBM suggested a nearby industrial sand/excavation site as another possible test location.

• Safety / Visibility / Transport Protection

  • KBM raised safety concerns for a large aircraft, including reflective/warning markings, motor/prop protection during transport, pin protection from dust/mud, and ground stability.
  • Julius has small 3D-printed feet on the corners to lift the vehicle slightly.
  • KBM suggested top/bottom strobe or position lights with red/green/white options.
  • Julius asked KBM to look for a practical light suggestion. Erick suggested battery-operated lights for PT1 so the team does not need extra wiring.
  • Erick mentioned Arrow has an unused parachute, likely in Turkey, that may work on 12 V and CAN. Julius treated parachute testing as more likely Phase 2; no decision was made to integrate it for the first hover.
  • Julius has been using trash bags over motors while drilling/working. KBM suggested more durable canvas/Oxford-cloth covers for protection during handling or transport.

• Quiver Side Update

  • Erick said another Quiver test looked good, Remote ID is figured out, GPS2 is working again, and ArduPilot/GPS calibrations are done.
  • Erick thinks the Quiver is ready for a test flight, but Julius advised against testing right before Erick leaves town.
  • Erick agreed he will likely wait until after returning Sunday rather than risk a crash before travel.

• Decisions / Direction

  • Caribou PCBs are ordered and waiting on manufacturing/shipping.
  • July remains the pressure window for the Phase 1 flight.
  • First flight prep should stay conservative: ground test first, then tethered hover-style flight.
  • Use simple off-the-shelf drone lights, likely battery-powered, for PT1 visibility instead of immediately wiring integrated lights.
  • Treat parachute integration as a possible later/emergency safety topic, not a confirmed first-hover requirement.

• Open Items

  • Julius: track PCB manufacturing and shipping status.
  • Julius: continue documentation cleanup and physical drone prep while PCBs are in production.
  • Julius: choose/confirm the first test site and transport route.
  • Julius: define the Phase 1 ground-test and tethered-hover sequence, including pass criteria.
  • KBM: find practical battery-operated or 12 V drone strobe/position light options.
  • Julius/KBM: add simple PT1 visibility/safety lighting after selecting a practical option.
  • KBM: think through additional large-vehicle safety/transport measures.
  • Julius: consider canvas/Oxford-cloth or similar covers for motor/prop/pin protection during handling and short trailer transport.

• Open Questions / Risks

  • Will PCB manufacturing/shipping hit the end-of-next-week target?
  • What exactly counts as the Phase 1 hover acceptance criterion?
  • Which local site is safe and practical for a tethered 200 kg-class prototype hover?
  • Does trailer transport need more formal protection for motors/props/pins?
  • Should the unused parachute be evaluated for first-hover emergency use, or deferred until Phase 2?