Federated Survey Data Sharing Between Expedition Teams

Three Teams, One Cave, Zero Merge Protocol

The Quintana Roo Speleological Survey was established in 1990 specifically as a federation model to coordinate multiple teams surveying overlapping Yucatán systems. Nearly 35 years later, the problem it was built to solve still shows up weekly: Team A finishes a 2-kilometer penetration in Sistema Ox Bel Ha, Team B surveys a parallel branch six months later starting from a different cenote entrance, Team C connects the two the following year. Without a federation protocol, each team's data lives in its own file structure, its own conventions, its own calibration history. Three partial quilts; no shared map.

The Dominican Republic Speleological Society's FEALC membership extends the same federation model to the Caribbean basin. Closer to the engineering side, Nature's foundational paper on FAIR Guiding Principles by Wilkinson et al. set the baseline expectations for scientific data federation: findable, accessible, interoperable, reusable. Nature's 2022 follow-up on FAIR in Earth science shows how community-centric metadata enable the kind of cross-team sharing that QRSS has been doing by convention for decades.

What goes wrong at the operational level: each team has a different compass correction, a different audio sample rate, a different survey-grade rubric. Merge-conflict resolution happens by email thread. Beyond the Sump expeditions — which depend on shared archives across multiple teams — document how much faster merged progress goes when federation is baked into the tool chain instead of patched on afterward.

The bandwidth cost of poor federation is real and measurable. A cross-team merge that runs through email threads consumes 40-80 hours of project-leader time per system per year, time that does not produce any new survey patches. NSS-CDS expedition leaders coordinating multi-team Florida cave projects have cited federation overhead as the single largest non-diving cost on their multi-year programs. WKPP and QRSS have built internal tooling over the decades to reduce that overhead, and EchoQuilt's federation protocol is intended to generalize those team-specific solutions into a common standard that any cave-survey federation can adopt without bespoke engineering work.

Stitching Federated Patches Into One Quilt

EchoQuilt's federation protocol works from one premise: each team keeps stewardship of its own patches, and the shared quilt is a merge view of what everyone has contributed. No one's data moves to a central server. The quilt is a federated graph.

Three building blocks make this operationally tractable:

1. Per-team patch provenance. Every patch carries the originating team's identifier, their survey-grade rubric version, and their calibration metadata. A Yucatán quilt stitched from QRSS contributions shows which team mapped which section, and a researcher querying the quilt can filter by team, by date, or by survey grade. Cave-diving literature documents that cave-survey data is collectively stored — the federation protocol makes that storage queryable.

2. Conflict flagging without conflict resolution. When Team A's patch disagrees with Team B's patch at a shared tie-point, EchoQuilt flags the conflict and presents both views. It does not auto-resolve. The federation's governance body (QRSS in Mexico, NSS-CDS in Florida) decides how to reconcile, and the audit trail records the decision. This is the same logic mine rescue coordination teams use for federated mutual-aid where different agencies hold partial workings maps — federation makes disagreement visible instead of hiding it.

3. Publishable quilt tiles. When a section of the federated quilt reaches publishable quality (survey-grade agreement across all contributing teams), it gets tagged as a publishable tile. Researchers outside the federation can cite the tile; the citation points back to each contributing team's patches. This links directly into the publishable maps audit workflow.

Three practical patterns that this enables:

Cross-basin joint expeditions. When a Mexican QRSS team and a Dominican DRSS team collaborate on a joint Caribbean karst survey, both teams' patches flow into a shared quilt without requiring either team to abandon their local data infrastructure.

Generational inheritance. Teams that dissolve (as they sometimes do, across 30-year spans) leave their patches in the federation. Successor teams can build on the orphaned data, which extends the expedition handoffs model into the federated domain. Cross-country data legality. Cave-survey data in some jurisdictions comes with legal constraints on dissemination. The federation protocol stores patches under each originating team's legal framework and shares only the attributes the team has authorized. No data movement, no legal ambiguity.

Federation Tactics for Multi-Team Projects

Projects coordinating more than three teams — QRSS, WKPP-adjacent mapping groups, multi-national karst consortia — run into three additional coordination layers:

Annual federation-wide calibration windows. Once a year, all contributing teams calibrate their hydrophones, compasses, and depth sensors against a shared reference standard at a designated cenote (usually Cenote Angelita for the Yucatán federation). The calibration results get published to the federation graph, and the subsequent year's patches get tagged against the calibration event. This keeps multi-team merge conflicts inside the measurement noise floor rather than drifting into systematic bias. Teams arriving outside the calibration window can still contribute patches but their data gets tagged as pre-calibration provisional until the next federation gathering retroactively brings them in line.

Survey-grade translation tables. QRSS, NSS-CDS, and European federations use different grading rubrics. EchoQuilt publishes translation tables so a team contributing at QRSS Grade 4 can map cleanly to BCRA Grade 5a for cross-federation publication. The translation is explicit, not implicit — a researcher always knows which rubric the original patch was recorded under. The translation tables are themselves federated artifacts maintained by the relevant federations rather than imposed from outside, which means changes to a rubric flow back into all dependent quilts the next time the federation publishes an updated mapping.

Federation-wide lead-hand-offs. A team that surveys past their gas-or-time envelope marks the far tie-off as a federation-shared lead. Any other member team can pick up the lead on their next expedition without coordination overhead beyond a pull-request-style patch contribution. This operational pattern is why QRSS-style federations cover more passage than any single team could map.

Permit and access coordination. Cave systems on private land, in protected areas, or on sites managed by indigenous communities require permits that take months to secure and that often restrict which teams may dive on which days. Federation protocols capture the permit holder, the access window, and any reporting obligations as part of the patch metadata, so a successor team picking up a lead knows whom to contact for re-entry permission. Sistema Sac Actun and the Wakulla Springs State Park dive program both require this kind of permit traceability, and the EchoQuilt federation graph stores the access metadata alongside the survey data so it cannot be lost when a team's institutional memory turns over.

Conflict-of-interest disclosure. When the same diver participates in two competing teams' survey efforts on adjacent or overlapping passages, the federation needs to know. EchoQuilt records the diver-by-diver contribution graph so a future audit can identify cases where one diver's patches contributed to two teams' competing claims on the same lead. The disclosure does not resolve the conflict, but it makes the conflict visible to the federation's governance body when it matters.

Federation turns multi-team cave survey from an email-thread problem into a graph-merge problem. The shift saves time, reduces ambiguity, and makes federation-scale survey work tractable for groups that today are bottlenecked on coordination rather than on diving capacity.

Build a Federation-Ready Survey Workflow

QRSS members, NSS-CDS project directors, and multi-country karst consortium leads coordinating three or more expedition teams need tooling that matches the complexity of the work. EchoQuilt's federation protocol lets each team keep stewardship while stitching patches into a shared quilt. Join the Waitlist for Cave Diving Survey Teams and talk to us about your federation's current archive format — first integrations prioritize federations with active cross-team expeditions in the next 12 months. Share your federation's contributing team count, your annual calibration window site (Cenote Angelita, Wakulla, or another reference cenote), your current grading rubric mix (QRSS Grade 4, BCRA Grade 5a, NSS-CDS publication standards, European federation grades), your permit and access-coordination structure (private land, protected area, indigenous community partner), and any existing federation-shared lead-handoff conventions.

We will scope a translation-table integration against your current rubric, prepare the federation-wide patch-contribution workflow with conflict-of-interest disclosure logging, set up the permit-traceability metadata schema so successor teams know whom to contact for re-entry, and configure the pull-request-style lead-handoff format. Priority access goes to QRSS, NSS-CDS, GUE, NACD, and IUCRR-affiliated federations with active cross-team campaigns scheduled.