Project: OpenSeis | The Distributed Fault Monitor

The premise of Project OpenSeis is simple but radical: seismic monitoring should not require a government budget or a five-figure server cluster. We want to build a global network of 'micro-stations' that anyone can replicate using open-source hardware and free cloud services. By sharing the schematic designs under the CERN Open Hardware Licence and releasing the firmware code on GitHub, we invite anyone—from hobbyist makers to university students in developing nations—to deploy a device in their backyard. The goal is to visualize the invisible, rhythmic grinding of the Earth's tectonic plates in real-time. We aren't trying to replace professional seismographs, but rather to fill in the blind spots between them, creating a 'noise correlation' dataset that is mathematically rich enough for scientists to study earthquake precursors without needing to own expensive proprietary equipment. It is a call to treat the Earth’s vibration as a public utility, accessible to all.

Seismology is notoriously expensive and opaque. High-quality seismic data requires specialized, power-hungry sensors and massive server infrastructure to process. Because of this cost barrier, the data is often siloed by government agencies or private universities, leaving vast swaths of the planet unmapped. Furthermore, the signals we care about—micro-foreshocks and slow slip events—are often localized and drown out in background noise. We lack a distributed, low-cost network capable of capturing these nuances and correlating them across distances. There is a massive gap between the scientific need for dense, continuous data and the technical barrier to obtaining it.

Hardware: ESP32-C3 microcontroller (low power), ADXL345 3-axis MEMS accelerometer, BMP280 barometer (for pressure compensation), SIM7600 GSM module (global connectivity), LiFePO4 18650 batteries (long cycle life), 3D-printed weatherproof enclosures (PLA+ material). Software: CircuitPython firmware for the sensors, MQTT over TCP protocol for data transmission, InfluxDB time-series database for raw data storage, Grafana for visualization, Python (Pandas/NumPy) for cross-correlation analysis scripts.

The project is built on a decentralized 'Node-Hub-Cloud' architecture. The Node is a battery-powered, cellular-connected device that samples analog acceleration at 100Hz. It performs local digital filtering to remove high-frequency noise (like wind) and sleeps between transmissions to save power. The node communicates via MQTT (Message Queuing Telemetry Transport) to a local gateway (a Raspberry Pi in a closet or a commercial MQTT broker). The Hub aggregates these streams and writes them to an InfluxDB time-series database. A Cloud Pipeline runs continuously, pulling data from the database and running cross-correlation algorithms on pairs of nodes to detect if the same vibration is felt simultaneously in two different locations. This signal is then pushed to the public Dashboard, allowing users to see if the ground moved right now.

The project started with genuine momentum—a successful Kickstarter and a community of 200 interested engineers—but it stalled when the lead designer took a full-time job in a different city and the regulatory paperwork for the GSM modules became a tedious bottleneck. There was also a sobering realization that maintaining the community infrastructure for a decade would require a full-time commitment that the volunteers simply didn't have. It remains a beautiful blueprint for how geoscience could be democratized, but it stayed exactly that: a blueprint.

This project was heavily inspired by the CUSP seismic network and the 'OpenSignals' project. The hardware design is intentionally modular to allow for future upgrades, such as adding soil moisture sensors or tilt-rotometers.