Project Quiver — February & March 2026 Progress Report

1. Executive Summary

In February and March, Project Quiver completed the transition from weekly contribution streams to a grants & bounties model while advancing towards a dev-kit launch.

February was defined by hardware troubleshooting and key technical decisions including a crash investigation, GNSS standardization on the F9P, and various hardware modifications. Documentation was fully migrated from HackMD to GitHub. Erick was confirmed as Project Lead and the March grants & bounties structure was scoped and prepared for governance.

March centered on documentation delivery and hardware validation. An information note bounty program drove seven notes to completion. Julius’ enclosure seal was validated in a water test and adopted as the primary design, and the transport case bounty was completed. The Tattu battery purchase was cancelled in favor of a community led custom battery project targeting a first prototype by end of April. QuiverHub received its first full team presentation and a Quiver Python SDK was proposed to complement it.


2. Project Progress

Team Formation

February

The Project Quiver team remained the same for February.

Member Experience Level Team Weekly Commitment (hrs) Areas of Expertise / Championed
21stCenturyAlex Level 3 Core 20 Avionics, QuiverHub, SDK
alperenag Level 4 Core 28 Project Co-Lead, Manufacturing Guide
Dow Fisher KBM Level 3 Core 15 FEA, Systems Engineering, PCB Design
errrks.eth Level 4 Core 35 Project Lead, Documentation, Bounty Administration
Julius Level 4 Core 35 PCB Layout, Power Storage, Prototyping, Enclosure Design, Electrical
ZeynepB Level 4 Core 30 Flight Mechanics, Software, Flight Tracking Platform
kjcerveny Level 3 Contributor 5 Electrical Design, System Testing, Product Development

March

In March, the project transitioned to a grants & bounties model. Contribution streams were reduced to commitment grants covering meetings, coordination, and general support for the dev-kit launch (~5 hours/week per contributor). The Project Lead receives additional compensation to cover administrative overhead including bounty administration, documentation review, payment processing, and governance responsibilities (~10 hours/week).

The following contributors received commitment grants for March:

Member Role
errrks.eth Project Lead
Julius Core Contributor
Zeynep Core Contributor
alperenag Core Contributor
Dow Fisher KBM Core Contributor
21stCenturyAlex Core Contributor

Progress Summary:

February opened with a crash investigation that ran across several team calls. The leading hypothesis was flight controller power loss caused by a compromised trace/connector failing under vibration leading to a loose connection mid flight. A pre-charge bypass MOSFET issue was identified as a contributing factor. Zeynep investigated whether a malformed CAN message could trigger a motor to disarm, covering the software failure path in parallel. In the end, the investigation yielded no single solution and the failure could not be replicated.

On the enclosure, a silicone O-ring and a liquid rubber seal were ordered for comparative waterproofing testing. Julius performed a water pour test on a printed enclosure and shared enclosure design updates with the team. In parallel, the attachment interface PCB was updated to use fewer parts, improving manufacturability and reducing cost by approximately $10 per board. PCB orientation was standardized and KiCad files were moved to their own repo folder.

The GNSS standard was settled after the Here4 GPS inexplicably failed to achieve RTK fix on one drone with no configuration or hardware explanation found. Thomas ordered eight F9P GPS units to standardize across all builds. The Here4 Blue was noted as a potential candidate for future Remote ID compliance. An RTK base station was simultaneously configured for Gray, a field operator actively conducting commercial survey work with Quiver, using an EMLID receiver cast over Starlink to enable sub centimeter positioning during operations.

Documentation moved off HackMD during this period. Erick completed the harnessing guide and opened the dev-kit branch on GitHub, establishing folder structure for all guides and image directories. Julius reviewed the harnessing model routing and provided feedback on cable paths. The nailboard manufacturing concept was noted as a future improvement for production builds. Zeynep began planning the dev kit user support structure, covering a Discourse forum channel, a pre-flight notification system, and automated log check criteria to support external operators after shipment.

Thinner carbon fiber tubes were approved for dev-kit builds. Thomas began batch producing motor beams, cutting and drilling 16 beams in preparation for the first kit assemblies. Early discussions on transport case options were initiated. Custom copper bus bars arrived for the power distribution system but were too short. Investigation found the battery connector PCB floating approximately 12mm too high in the Fusion 360 CAD assembly, a model positioning error that had gone unnoticed after a software update. OSH Cut was identified as a source for custom cut and bent replacements, with its integrated tube bending workspace demonstrated live on a call. Julius also explored ordering from a German supplier. Replacement bus bars were subsequently ordered and confirmed received.

Late February included a structured review of Julius’ enclosure design updates. The changes covered antenna cones beneath the mounting surfaces to improve drainage, an inclined LiDAR surface, and silicone sealing between aluminum layers to replace the LiDAR cable grommets, which were too small for the cable opening. The front cable tray was dropped in favor of routing cables in from the side, and M3 insert holes were deepened to properly seat the McMaster Carr inserts. The battery connector area was flagged for a water pour test with the battery installed and latched to verify the rubber seal holds under rotor wash. Both seal candidates were available in comparable dimensions on Amazon and Thomas ordered both for comparison, with the goal of writing a formal seal specification. The commits were recommended for merging as refinements that were not flight critical, excluding the LiDAR grommet change.

Documentation integration was nearing completion. The assembly guide was largely finished, pending image sizing and final editorial work, with Julius’ updated enclosure to be integrated by end of week. Erick and Alperen planned to merge the PCB and harnessing sections into the manufacturing guide once that was in. A concern was raised about HackMD image hosting. Information notes that had been migrated to GitHub still linked to HackMD hosted images, which raised questions about long term reliability.

Zeynep got the flight report generator working with help from Vector. The tool uploads a log file and generates a consolidated view of key flight data comparable to manual log analysis. A GPS/4G RF interference issue was confirmed during a flight attempt where the Here4 GPS acquired no satellites while the 4G module was active on the Raspberry Pi. Removing the module restored signal. Julius planned further testing to confirm before the 4G module was ruled out. A consolidated autopilot firmware was uploaded to GitHub merging the Ethernet PPP configuration and required LiDAR changes. QuiverHub’s Tailscale integration was confirmed working, with Julius successfully accessing Mission Planner over LTE via the Raspberry Pi.

The month closed with the grants & bounties structure reviewed by the full team. The total budget was confirmed to flow into the Quiver project wallet with Erick as administrator. Wind limit values were queued for update in the pilot manual for both automated and manual flight profiles, and the obstacle avoidance bounty write up was assigned.

March opened with a full team review of the grants & bounties structure and began payouts as deliverables were completed. Julius’ gasket based enclosure seal was validated in a water test with no water intrusion observed, and was adopted as the primary assembly design. KBM’s enclosure and sensor mount information notes document both the gasket and liquid silicone approaches for reference.

The transport case bounty was completed by Julius, who sourced a used Pelican 1640 case and cut three layer custom foam inserts using a diode laser cutter. The drone fits with the LiDAR top cap installed and the case closes fully. Alperen designed a complementary camera cover, which Julius printed and tested. The team shifted preference away from the Nanuk case (up to €1,300 in Europe) to the Pelican platform.

An information note bounty program drove completion of seven notes, offering a 300 $ARROW bonus per note committed to GitHub by March 11. $ARROW tokens were paid out and a liquidity pool was seeded on Ethereum L1, giving contributors a path to convert $ARROW to USDC.

Design sessions for the Universal Cargo Container produced two bounties: (1) a V1 actuated latch/release mechanism (servo + hook latch, powered from the drone’s 12V switched output) and (2) a cargo container/enclosure. Requirements: max 6kg payload, max ~1kg container weight, cylindrical preferred shape, footprint not to exceed drone body.

The Tattu 33Ah battery purchase was cancelled in favor of a custom battery project led by Julius. The pack targets the same form factor as the Tattu while delivering approximately 45% more energy using 126 high capacity 21700 Li-ion cells. The Li-ion chemistry also offers a better safety profile and longer cycle life than LiPo for commercial field use. The project runs in two phases. Phase one targets a functional prototype by end of April covering mechanical design, cell assembly, and bench testing. Phase two adds a full smart BMS with protection, balancing, and CAN communication. KBM’s Tattu adapter PCB will be retroactively paid and added to the repo as a reference design.

QuiverHub was formally presented on March 23. Key capabilities include:

  • Flight State app pulling telemetry via MAVSDK-Python
  • Camera app with live CE camera streaming
  • Flight Analytics app with interactive log viewing, GPS map overlay, and multi log comparison
  • OTA and Log Capture for wireless firmware updates and log downloads
  • Mission Planning for waypoint upload and autonomous mission execution
  • Multi drone fleet support
  • App Builder with drag and drop UI construction

Thomas proposed a Quiver Python SDK to provide simple high level drone control abstractions on the Raspberry Pi, complementing QuiverHub’s remote operator interface. Alex created a quiver-sdk repo with an initial SPEC.md and a discussion thread to gather input from the team.

Documentation progress in March included: Pilot’s Handbook formatting fixes (GitHub compatible warning and note block rendering), near completion of the Developer SDK Guide (quick start guide remaining), manufacturing guide chapter ordering and image link improvements, and ongoing assembly guide integration work by Erick and Alperen. All documentation images were moved to GitHub hosting, resolving the HackMD reliability concern raised in February. The LiDAR cable grommet issue flagged in February was resolved. Wind limit values were added to the Pilot’s Handbook for both automated and manual flight profiles.

A drone identification design session produced a near term plan: automotive grade stickers covering serial number, Arrow logo, and a QR code linking to the repo, with a “founder series” framing for the first batch. A laser engraving bounty was discussed as a longer term option and Gavin is working on the design.

Remote ID configuration remains a pending blocker before the first dev kits can ship to Ben and Boosh. Thomas completed calibration on his build but the Raspberry Pi integration and Remote ID verification are still outstanding.

The final week of March saw reduced output across the project. Julius was traveling from March 19 to 27, and the project lead was out sick in the same window, which accounts for several items carrying into April.


3. Information Notes

The following information notes were completed and committed to GitHub during February and March:


4. Goals for Next Month

  • Complete and publish all remaining dev-kit documentation (assembly guide, manufacturing guide, engineering report).
  • Add a quick start guide to the Developer SDK.
  • Proceed with QuiverHub grant.
  • Scope and assign Quiver Python SDK bounty.
  • Complete V1 actuated latch/release mechanism bounty.
  • Conduct endurance metrics flight testing across payload and flight profile matrix.
  • Begin obstacle avoidance flight testing (pending Top360 LiDAR resolution).
  • Ship first dev kits to Ben and Boosh and begin external field testing.
  • Evaluate and resolve attachment interface supply risk (manufacturer ending production).
  • Resolve GPS/4G RF interference.

5. Budget & Resource Allocation

Project Expenses:
$1,299 was reimbursed over the course of February and March for Fusion 360 tokens. Expenses are tracked in the reimbursements sheet

Team Members Compensation:
The project team was compensated via contribution streams in February. For March, the following commitment grants and bounties were allocated:

March 2026 — Total Budget: $35,000 USDC + 15,000 $ARROW

Category Budget
Commitment Grants $8,760 USDC
Grants & Bounties $26,240 USDC
Retroactive Grants 15,000 $ARROW

Commitment Grants — $8,760 USDC

Contributor Grant (USDC)
Erick $3,000
Julius $1,280
Zeynep $1,280
Alperen $1,280
KBM $960
Alex $960

Grants & Bounties — $26,240 USDC

ID Name Reward Status
QGB-01 Documentation Wrap-Up $10,000 :yellow_circle: In Progress — $1,764 USDC paid. $8,236 USDC remaining
QGB-02 Obstacle Avoidance System $4,000 :white_circle: Open — blocked pending Top360 LiDAR resolution in Texas
QGB-03 Endurance Metrics Study $600 :white_circle: Open
QGB-04 Wind Limit Study $3,000 :white_circle: Open — deliverable repurposed: wind tracking to be integrated as a permanent flight tracking app feature rather than a one-time study. Zeynep to analyze existing flight data to establish initial wind range
QGB-05 Attachment Development $3,000 :white_circle: Open — design sessions completed. Cargo container bounty TBD
QGB-05a V1 Actuated Payload Latch $500 :white_circle: Open — sub-bounty from QGB-05. Milestone-based ($150 / $150 / $200)
QGB-06 Transport Case $1,000 :yellow_circle: Pending payment — physical build complete (Julius, Pelican 1640 + laser cut foam inserts). Awaiting information note to close bounty
QGB-FLEX Multispectral Camera Payload $500 of $4,640 :white_circle: Open — MAPIR Survey3W RGN mount. Milestone-based ($150 / $150 / $200)
QGB-FLEX Fusion 360 Tokens (March) $649.50 of $4,640 :white_check_mark: Paid — $2,991 remaining in flex bucket

QGB-01 Breakdown

Item Allocation Status
Dev-Kit Information Notes $775 USDC :white_check_mark: Paid
Liquidity Pool Seeding $989 USDC :white_check_mark: Paid
Manufacturing Guide $2,000 USDC :white_circle: Open
Engineering Report $1,000 USDC :white_circle: Open
Pilot’s Handbook $1,000 USDC :white_circle: Open
Developer SDK Unpriced :white_circle: Open
Maintenance Guide Unpriced :white_circle: Open
Remaining (USDC) $4,236 USDC After priced items

Retroactive Grants — 15,000 $ARROW

Contributor Amount
Total Reserve 15,000 $ARROW
Info Note Bonuses 3,300 $ARROW
Liquidity Pool Seeding 3,300 $ARROW
Erick Retroactive Grant 3,000 $ARROW
Julius Retroactive Grant 1,500 $ARROW
KBM Retroactive Grant 1,000 $ARROW
Thomas Retroactive Grant 1,000 $ARROW
Alperen Retroactive Grant 800 $ARROW
Zeynep Retroactive Grant 450 $ARROW
Alex Retroactive Grant 300 $ARROW
Sleety Retroactive Grant 200 $ARROW
Ben Retroactive Grant 150 $ARROW
Remaining 0 $ARROW

Total:

The total expense of Project Quiver in February was $37,052.50 ($36,403 in commitment grants + $649.50 Fusion 360 token reimbursement), which is below the monthly maximum spending cap. In addition, team members received 20,224 $ARROW as part of their compensation.

For March, the total budget is $35,000 USDC + 15,000 $ARROW. Paid to date: $11,173.50 USDC (commitment grants $8,760 + QGB-01 payouts $1,764 + Fusion tokens $649.50) and 15,000 $ARROW (info note bonuses + liquidity pool seeding + retroactive grants). An additional $1,000 USDC is pending payout for QGB-06 once the transport case information note is complete. The remaining budget is held for open bounties.

2 Likes