Wandering With Pride · Miami pilot 2026

Safety that works when nothing else does.

WanderSafeMesh is a community-owned emergency mesh in development — 100 solar nodes designed to deliver safety information, emergency alerts, and crisis contacts to LGBTQ+, Deaf, disabled, undocumented, and unhoused residents when apps, cell towers, and power grids fail. Miami pilot launching 2026.

Open interactive 3D model →
Node v1 · 100 × 60 × 170 mm — fits in your hand · mounted on a Little Free Library post · 3D-printed white-ASA IP67 · multi-face thin-film solar · LoRa 915 MHz · LiFePO4 6 Ah

All images are concept renders of the planned hardware. Physical units are not yet built — Miami pilot launches 2026.

100 nodesMiami pilot 2026
$144 / nodeOpen-source BOM
72h offlineLiFePO4 6Ah + thin-film solar
Zero surveillanceArchitecturally enforced

The gap

Apps fail when emergencies are real.

During Hurricane Ian, Miami-Dade's cell towers overloaded in hours. Hurricane Sandy knocked out 25% of U.S. cell towers. The IPCC projects extreme weather events will double in frequency this decade. Miami-Dade already has 250,000 residents living below 3 feet of sea level, with 21–54 inches of rise projected by 2070 — making offline resilience infrastructure not a backup, but a necessity.

During an ICE sweep, a cell-connected app is a liability. When you're Deaf, unhoused, or don't speak English, official emergency systems were never designed for you. WanderSafeMesh is.

The mesh

Offline by design. Community by architecture.

Each node is designed to run AES-256 encrypted LoRa mesh firmware — connecting to neighbors without internet, cell service, or power. Icon-first e-ink display for Deaf users. Audio mode for blind users. Three tactile buttons. Panic button routes to community responders — never police.

  • No internet, no problem. 915 MHz LoRa radio reaches ~1.2 km in dense urban terrain. The pilot will deploy 100 nodes across 10 Miami-Dade neighborhoods.
  • No cell tower needed. Mesh self-heals around any failed node. Hurricane outages don't cripple the network.
  • No login, no account. Walk past an LFL post, read the icons, that's it.
  • No location tracking. Nodes broadcast environmental data, not visitor data.
  • Three languages on every display. English, Spanish, Haitian Creole — chosen at the city-config layer.
Miami neighborhood map with WanderSafeMesh nodes in Wynwood, South Beach, Overtown, Liberty City, and Little Havana, connected by mesh signal lines.
Phase 1 · 10 neighborhoods · 100 nodes · ~1.2 km LoRa range

The hardware

$144. Solar. 72 hours. Open source.

3D-printed white-ASA enclosure (thermal-protected for Florida sun). Heltec WiFi LoRa32 V4 (ESP32-S3 + SX1262, 915 MHz). LiFePO4 6 Ah battery. Multi-face thin-film solar (~4 W total across top + both sides). IP67 weatherproof. Designed to mount at ADA height on Little Free Libraries across Wynwood, South Beach, Overtown, Liberty City, and Little Havana.

Exploded technical illustration of the WanderSafeMesh node showing top thin-film solar panel, PCB with ESP32-S3 + SX1262 LoRa module, LiFePO4 battery cell, and four sensors (BME680 atmospheric, PMS5003 air quality, HC-SR04 ultrasonic flood, AS3935 lightning detection).
Exploded view · top thin-film solar · ESP32-S3 + SX1262 main PCB · 6Ah LiFePO4 · 4 sensors on bottom PCB · IP67 ASA shell · ~$144 BOM
ComponentSpec
RadioLoRa 915 MHz · AES-256 mesh · ~1.2 km urban range
MCUHeltec WiFi LoRa32 V4 · ESP32-S3 · SX1262
BatteryLiFePO4 6 Ah · 72h offline · 330-day reserve
SolarMulti-face thin-film ~4 W · top + both sides · orientation-agnostic
Display2.13″ e-ink (GDEH0213B72) · sunlight-readable · 0 mW standby
Interface3 nav buttons + panic button (flip-guard) · passive buzzer audio
SensorsBME680 · PMS5003 · HC-SR04 · AS3935
EnclosureIP67 · white ASA (UV + thermal) · 3D-printed · STL open source
SecurityESP32-S3 secure boot v2 · flash encryption · panic-button rate limit
Cost$144 / node · CC BY 4.0 firmware
Front face: e-ink display, 3 nav buttons, panic button with flip-guard, WanderSafe wordmark.
Front
Back face: smooth white ASA with stainless steel pole-clip mount bracket.
Back
Left side: full thin-film solar panel laminated flush.
Left
Right side: solar panel and visible stainless steel pole-clip mount bolts.
Right
Top: thin-film solar panel covering full top face.
Top
Bottom: 3 IP67 sensor ports, USB-C charging port with rubber cap, IP67 membrane vent.
Bottom

The sensor suite

Four sensors. Each one earns its place.

Every sensor must serve a directly observable community need. Microphones, cameras, and passive Bluetooth/WiFi scanners are permanently off the component list — not as policy, as a hardware bill of materials choice.

BME680 · I²C

Heat · humidity · pressure · VOC

±0.5 °C, ±3 % RH, ±1 hPa, plus indoor-air-quality VOC sensing. A pressure drop above 3 hPa in three hours triggers a storm advisory. Heat index above community threshold triggers a heat alert — the same metric NIOSH uses for outdoor worker exposure. The barometric channel also gates flood alerts (no false alarms during routine Miami rain).

PMS5003 · UART

PM2.5 · PM10 air quality

Designed for a continuous PM2.5 grid across five Miami neighborhoods once the pilot is deployed. EPA EJScreen calibration candidate. Thresholds set so PM2.5 above 35 µg/m³ will trigger an advisory; above 55 µg/m³ an unhealthy alert. Vented PTFE-membrane port keeps rain out, air in.

HC-SR04 · Ultrasonic

Flood · water level

Distance-to-water at 5 cm accuracy. Above 50 mm baseline triggers a flood advisory; above 200 mm triggers an evacuation alert. Privacy: only neighborhood-aggregate flood data ever leaves the city — node-level levels could reveal who's sheltering in place.

AS3935 · Franklin SPI

Lightning detection

1–40 km range, 14-step distance algorithm. With three nodes detecting the same strike, TDOA analysis is designed to localize within ~1–2 km — building toward a community-owned urban lightning map. Lightning is the #1 cause of weather-related outdoor deaths in Florida — Pride, Calle Ocho, and beach events sit in the bullseye.

The covenant

Zero surveillance. Architecturally enforced.

Trust isn't a policy. It's a hardware constraint. A community member who walks past a node, has no phone, speaks no English, and has never heard of WanderSafe should be able to confirm the device isn't surveillance — within 30 seconds, without help from us.

  • No camera. No microphone. No Bluetooth or WiFi scanning. No GPS lookup of devices passing by.
  • AES-256 encrypted LoRa mesh — community holds the keys, not WanderSafe staff.
  • Every node broadcasts a signed null-surveillance declaration every 15 minutes, readable by any nearby device.
  • Firmware is CC BY 4.0 open source. Anyone can verify what the node does and does not do.
  • Warrant canary built into firmware. Any compelled disclosure breaks the canary — making government pressure architecturally visible.
  • A community board — elected by the neighborhood, not WanderSafe — holds veto authority over all researcher data access. Organizational leadership cannot override it.
  • STL files are CC BY 4.0. Any makerspace can print replacement parts.
  • Government cannot subpoena data that does not exist. Architecture, not promise.

The 10-year roadmap

Miami is the pilot. 100 cities is the goal.

Miami isn't the destination — it's where we build and prove the model. Once the pilot demonstrates the architecture in Miami's flood corridors, ICE-vulnerable neighborhoods, and queer community network, what emerges is a replicable, open-source, community-sovereign resilience infrastructure that any marginalized-population organization can deploy in any city.

A WanderSafeMesh node mounted on a white-painted Little Free Library wooden post in a residential Miami neighborhood, with daylight, blurred suburban background, and the device's solar panel and stainless steel mounting bracket clearly visible.
Designed deployment context · Little Free Library host post · ADA mount height
Phase 1 · Year 1–2

Miami

Target: 100 nodes · 10 neighborhoods · 15 LFL hosts. Funding pursuit: Knight Cities Challenge ($200k application). Partner discussions in progress with SAVE Miami and other LGBTQ+ orgs.

Phase 2 · Year 2–3

Knight 4-city

Planned expansion: Detroit, Philadelphia, Long Beach, New Orleans. ~400 additional nodes. FEMA BRIC application path identified.

Phase 3 · Year 3–5

Metro 8

Target metros: NYC, LA, Chicago, SF, Seattle, Austin, Portland, Atlanta. ~2,200 nodes. Funding paths identified: NSF CIVIC, HRSA Health Equity. First peer-reviewed paper as a Phase-3 milestone.

Phase 4 · Year 5–7

Broaden

+16 additional cities. Indigenous communities, farmworker regions, public housing, reentry communities. Same architecture, city-config adapted.

Phase 5 · Year 7–10

National vision

Long-term target: 100 cities · ~10,000 nodes · ~1,500 paid community ambassadors — a first-of-its-kind community-owned national environmental safety network.

Open by design

Build your own. Join the network.

Miami is the pilot, not the limit. The hardware spec, firmware, and STL files are planned for CC BY 4.0 release — so anyone, anywhere, can build a node, run their own neighborhood mesh, and (when authentication ships) bridge into a wider community of safety nodes. Even one node on a single block helps that block. Density compounds: there's no minimum threshold to start.

Replicate

Build your own node

~$144 in off-the-shelf parts. No special tooling beyond an FDM 3D printer and a soldering iron. The full bill of materials, STL files for the IP67 enclosure, and firmware are planned for CC BY 4.0 release after the pilot firmware lands. Documentation aimed at any community group — not just engineers — with build cards in English, Spanish, and Haitian Creole.

Integrate

Existing networks we build on

WanderSafeMesh sits on top of Meshtastic — the global open-source LoRa mesh community. Governance is informed by NYC Mesh, Toronto Mesh, and Freifunk. Where infrastructure permits, gateway nodes can bridge to The Things Network for wider LoRaWAN reach without locking the local mesh into any one ecosystem.

Authenticate

Trust without a central authority

The hard problem with "anyone can build a node" is rogue nodes. The planned trust model: cryptographic device attestation at mesh join (ESP32-S3 secure boot v2 + flash encryption), node whitelisting for critical-safety channels, community-reported revocation propagated mesh-wide, and tiered access — local mesh open to anyone, critical safety channels require attested devices. Anyone can build; the network maintains integrity.

Compound

Density wins

A solo node on a single block still helps that block. Two nodes within ~1.2 km mesh together. Ten nodes in a neighborhood become resilient against any single-node failure. The architecture scales bottom-up: every additional builder, every additional block, every additional city makes the coverage stronger and the network more useful. No minimum threshold to start. No central permission to ask.

Spec releases, firmware, and STL files will be tracked at github.com/WanderingWithPride once the pilot firmware is ready.

Common questions

If you're new to mesh networks.

Who does WanderSafeMesh actually serve?
The communities existing emergency infrastructure fails: undocumented immigrants, unhoused people, LGBTQ+ residents with histories of institutional harm, low-income renters in flood-prone neighborhoods, and environmental-justice communities with disproportionate air quality and climate risk. The network exists to serve them — not data brokers, not governments, not researchers.
Why not just use cell phone alerts?
Existing tools have four gaps: they require internet and smartphones; they require trust in institutions (which many communities don't have for good reason); they don't monitor hyperlocal hazards (NWS data is regional, the corner you live on is not); and they don't feed data back to the community. WanderSafeMesh is designed to close all four. When the pilot is deployed: walk past a node, see the display, that's it.
Is the data sold?
No. Aggregate environmental data is published CC BY 4.0 — anyone, including researchers, can use it. Cost-recovery revenue from research subscriptions and event APIs funds maintenance. The hard rule: revenue is cost-recovery for infrastructure, never profit extraction from the community.
What happens if WanderSafe (the org) shuts down?
Open-source firmware (CC BY 4.0) and open-source STL files mean any makerspace can keep nodes running. Each city's QR code resolves to a community-partner-hosted page (SAVE Miami discussions in progress for Phase 1), not WanderSafe's server. The local community board can keep operating without us.
Can I build my own node and deploy it somewhere other than Miami?
That's the design intent. The hardware spec, firmware, and STL files are planned for CC BY 4.0 release after the pilot firmware lands, so any community group anywhere can fork the build, run their own neighborhood mesh, and (when the authentication layer ships) bridge into the wider WanderSafeMesh-compatible network. Even one node on a single block helps that block — coverage compounds with density, and there's no minimum threshold to start.
How do you keep rogue nodes from joining once it's open worldwide?
The planned trust model is layered: ESP32-S3 secure boot v2 + flash encryption (firmware integrity), cryptographic device attestation at mesh join (verifies the node before it relays), node whitelisting for critical-safety channels (panic-button traffic stays inside attested nodes), and community-reported revocation propagated mesh-wide (a compromised node gets quarantined fast). Local mesh participation stays open to anyone. Critical safety channels require attestation. Anyone can build; the network keeps its integrity.
Why Little Free Libraries?
LFL stewards are community trust anchors — already part of neighborhood fabric, maintained by volunteers who live there. LFL hosts get a physical education kit, a laminated card explaining the device, and direct contact with the community board. The node on an LFL works for everyone: no login, no phone, no English required.
How does the panic button route?
To a trained community responder — never police. In the Miami pilot, the planned routing is to ambassadors trained by partner LGBTQ+ orgs (SAVE Miami discussions in progress) who complete a 4-hour trauma-informed triage module before going live. Routing targets: Trevor Project or Trans Lifeline for mental health crises, immigration legal aid for ICE-related alerts, domestic violence hotlines for safety threats, and 911 only for life-threatening emergencies — and only at the responder's discretion. The architecture never auto-dispatches — every alert routes through a trained human. The no-police rule is enforced in firmware, not policy.

Three doors

Pick the one that fits.

WanderSafeMesh is funded by Wandering With Pride Inc. (501(c)(3), EIN 99-3467744). Contributions to support hardware deployment are tax-deductible.

For partner organizations

Partner with us

SAVE Miami, LFL stewards, neighborhood ambassadors, city emergency management. We fund the hardware; you anchor the community trust.

Partner inquiry →
For foundations & grant officers

Fund a phase

Phase 1 (~$200k target) funds the Miami pilot. Phase 2 (~$1.5M target) is the proposed four-city Knight expansion. Phase 3 has a FEMA BRIC funding pathway identified. Diligence packet on request.

Funder inquiry →
For volunteers & ambassadors

Help build it

Solder a node. Print an enclosure. Translate a display string. Steward an LFL. The hardware is open source and the work is real.

Volunteer →
Or make a tax-deductible gift →