Community Stoat Chat
Live Network v1.0 Beta

Early Warning
Seismic Network

Detect the P-Wave before the destructive S-Wave arrives. Get precious seconds to take action with our community-powered sensor network.

MQTT Protocol
ESP32 + LSM6DSO
Open Source

208 Hz Sampling

High-frequency detection for accurate P-wave identification

Multi-Node Validation

Cross-correlate signals to filter noise from real events

Seconds Matter

Get precious time to take protective action

Catch the P-Wave

Early Warning in Seconds

Our system works in 3 simple steps to catch the P-Wave:

1

Sensor Detects

Your GeoShake sensor detects the first, harmless P-Wave of the earthquake in a fraction of a second and instantly reports to the GeoShake Network.

< 0.5s
2

Network Validates

Our central system compares data from thousands of sensors in real-time. If 3+ sensors report signals simultaneously, it confirms as a "Real Earthquake".

Multi-Validation
3

Phone Alerts

The verified alarm is sent to your phone via the GeoShake App as an instant notification seconds before the destructive S-Wave reaches you.

Instant Alert

Open Hardware, Authorized Data

GeoShake's unique structure democratizes hardware while implementing strict security measures to maintain scientific data quality.

Fully Open Hardware

Schematics, PCB designs, and firmware code are open under MIT license. Buy ready-made or build your own.

  • ESP32 / ESP8266 Support
  • LSM6DSO Sensors
  • DIY and Ready Device Support

Authorized Cloud Network

To prevent data pollution, only registered and authenticated devices can write to the central database.

  • Unique Device ID & Key
  • Secure MQTT (TLS) Connection
  • Data Security with Supabase RLS

System Architecture

Three-layer, scalable, and modern structure.

1. Edge Device

Samples sensor data, performs local filtering (PGA/RMS), and packages only important summary data.

100Hz Sampling

2. MQTT & Ingest

Broker receives incoming packets, Ingest service validates and writes to database. Provides bidirectional control.

Pub/Sub Model

3. Supabase Cloud

Data is stored and analyzed. Anomalies are scanned via Edge Functions and served via API.

PostgreSQL + Realtime
payload_example.json
// Topic: geolink/v1/GLK-T1-847/telemetry
{
  "schema": "geolink.telemetry.v1",
  "device_model": "GeoShake T1",
  "sent_at": 1731865200123,
  "window_sec": 30,
  "samples": [
    { 
      "t_offset": 0, 
      "pga_g": 0.0123, 
      "rms_g": 0.0045, 
      "event_flag": false 
    },
    { 
      "t_offset": 1, 
      "pga_g": 0.0342, // High acceleration detected
      "rms_g": 0.0101, 
      "event_flag": true 
    }
  ],
  "sensor": { "type": "LSM6DSO", "axis": "Z" }
}

How Can You Join the Network?

Get a plug-and-play device, or build your own with open source schematics.

RECOMMENDED

GeoShake T1

Plug & Play Seismic Station

GeoShake T1 device
2-Min Setup WiFi Ready TLS Encrypted

Processor

ESP32 WROOM

Sensors

4× LSM6DSO

Sample Rate

208 Hz

Range

±2g

Advanced Analysis

PGA Detection P/S Wave Analysis Arias Intensity Offline Buffer

Mobile App

Coming Soon

View Details

DIY Edition

Build Your Own

For electronics enthusiasts. Download PCB files, source the parts, and build your own seismic station. Complete freedom.

Supported Boards ESP32 / ESP8266 / Pi
Sensor Options LSM6DSO/LSM6DS3
Firmware Open Source (GitHub)
Cost ~$10 - $25 (Estimated)

Why GeoShake?

Network Density

Instead of expensive and few sensors, we increase spatial resolution in P-wave detection by using thousands of low-cost sensors.

Community Power

Every user is a data provider. The community, not a central institution, handles system maintenance and growth.

Speed Focused

Instead of heavy seismic data packets, we reduce analysis time to milliseconds by sending "Event Flag" with lightweight MQTT messages.

Critical Questions & Project Logic

Why cheap sensors instead of expensive and precise seismometers?

For early warning (P-wave detection), Network Density is more important than individual sensor sensitivity.

Instead of a single $10,000 geophone, using thousands of $20 LSM6DSO sensors to build a 500x denser network is the only sustainable way to detect the epicenter faster.

How is it different from Raspberry Shake?

Raspberry Shake is a scientific archive-focused station with closed hardware/software that records full waveforms with high precision.

GeoLink is designed for early warning, open source and low cost; it captures the P-wave and validates by correlating thousands of sensors, sending lightweight MQTT event data instead of heavy waveforms.

In summary, Raspberry Shake is a single, high-accuracy station; GeoLink is an experimental early warning approach based on many cheap stations for speed and coverage advantage. They complement each other, not replace.

How do you distinguish between noise (Truck, Door Slam) and an earthquake?

This is the project's biggest technical challenge. Cheap sensors alone are very susceptible to noise. Our solution is the Correlation Algorithm.

If only your device is shaking, it's noise (door slam, passing truck). However, if 10 different devices within a 5 km radius give similar signals at the same time (within milliseconds), it's a seismic wave.

What about device placement and user responsibility?

Mounting the device on a fixed surface (preferably a column or beam) is critical. A device on a wobbly table or soft carpet is useless.

Also, an Early Warning System must run 24/7. Users not unplugging the device and not disconnecting internet is vital for system health.

Why is MQTT protocol used?

Traditional seismometer networks (SeedLink) carry full waveforms for scientific archiving and are heavy. GeoLink is Speed Focused.

With MQTT, we only send the "EVENT DETECTED!" flag and summary data (PGA/RMS). This is the most sensible solution in scenarios where even a thousandth of a second matters for P-wave detection.

Is there a risk of false alarms? What's the legal status?

Yes, there is risk. That's why GeoLink is an experimental community project and does not replace official earthquake data from authorities.

To minimize false alarms, our algorithms require high threshold values and multi-device consensus.

How much electricity does the device use? Will it affect my bill?

GeoLink devices (ESP8266/ESP32) consume extremely low energy. Average power consumption is around 1-2 Watts.

This consumption is much lower than a modem or LED bulb. Even with 24/7 operation, the impact on your electricity bill is negligible per month.

Will it slow down my internet? How much data does it use?

Absolutely not. Thanks to MQTT protocol, the data sent are very small text files. You won't notice any slowdown while watching videos or playing games.

Even if the device sends data every 5 seconds while idle, since the packet headers are very small, the average monthly total data usage stays around 15-30 MB. Data-cap friendly.