Research pilot · Open development

Open instruments.
Complete data.
Resilient field science.

A research-first particulate matter sensing platform designed to keep measuring through network failures, preserve every accessible sensor output, and make trustworthy air-quality research easier to share.

Measurements are for research and education—not regulatory decisions or health alerts.

FIELD NODE SPS30 Research
Sampling
PM outputs10
Local recoverySD
PublishingPrivate
01

Preserve the signal

Retain the complete accessible sensor output—not only a dashboard PM₂.₅ value.

02

Expect disconnection

Keep sampling locally and recover uploads without logical duplicates.

03

Publish deliberately

Devices begin private. Public locations and data require separate approval.

The platform

From field observation to usable scientific record

The everyday experience stays simple while the system handles provenance, offline recovery, authentication, and privacy in the background.

01

Measure locally

ESP32-S3 nodes collect complete SPS30 output together with environmental, timing, firmware, and quality metadata.

  • Mass and number concentrations
  • Temperature, humidity, and diagnostics
  • Persistent SD buffering
02

Store reliably

At-least-once delivery and stable observation identifiers allow safe retry after Wi-Fi or server interruptions.

  • Accepted means committed
  • Idempotent batch ingestion
  • Immutable raw observations
03

Share with context

Approved deployments can publish carefully selected data, location precision, attribution, and quality information.

  • Private by default
  • Deployment-specific consent
  • Public API field allowlist

Research direction

More than a low-cost concentration reading

Optical PM sensors respond to particle size, humidity, composition, optical properties, and instrument state. Open PM Network preserves the inputs needed to study those influences rather than hiding them behind a single number.

Current boundary

Low-cost measurements are not substitutes for regulatory or research-grade reference methods. Corrections must be validated against clearly defined, collocated reference measurements.

01

Complete output retention

Keep manufacturer-accessible measurements, status fields, and unknown future schema fields for reprocessing.

02

Physics-informed calibration

Evaluate how size, humidity, hygroscopicity, and optical properties influence sensor response.

03

Uncertainty-aware QA/QC

Pair corrected values with quality flags, calibration domain checks, and sensor-health evidence.

04

Event-responsive sampling

Explore plume detection and autonomous control of filter sampling for subsequent analysis.

21 days

of preliminary outdoor field comparison completed in the Atlanta region

Early field work

Encouraging consistency, carefully described

The first outdoor SPS30 deployment achieved 97.3% coverage of eligible hourly bins and followed major variations seen across nearby sensor networks. Because the instruments were not collocated, this is a field consistency check—not a calibration or accuracy validation.

Read the public field report

Community and classrooms

Simple participation without giving up data governance

The first community workflow is designed around three actions: add a sensor, view or download data, and choose whether to make a deployment public.

Teacher-managed pilots

School staff manage devices. Students can explore approved classroom data without creating personal accounts.

Location privacy

Private and public coordinates are separate. Owners choose the public location precision and preview it before approval.

Responsible openness

Public data includes attribution, units, quality context, and a clear distinction between access approval and scientific certification.

Roadmap

Start small.
Build evidence.
Open carefully.

  1. Phase 1
    Research team pilot

    Validate complete SPS30 data capture, offline recovery, private uploads, and deployment history with 2–20 nodes.

  2. Phase 2
    K–12 pilot

    Add teacher-managed deployment, simple classroom views, downloads, and school-level attribution.

  3. Phase 3
    Public community network

    Expand self-service registration, public APIs, integrations, and governance when real demand appears.

Data portal

Private pilot access is being prepared.

The sensor portal will open at pm.openpmnetwork.org after authentication, ingestion, and privacy checks are complete.

Portal coming soon