V.O.I.D. is an edge-to-cloud telemetry pipeline and Delay-Tolerant Network (DTN) engineered for emergency disaster scenarios. When standard communication grids fail, the system utilizes low-power microcontrollers to autonomously form a survival mesh network, buffering telemetry and bridging it into a cloud-native Kubernetes observability stack upon infrastructure restoration.
- Hardware Agnostic: Firmware compiles and executes dynamically across both ESP32 (Core v3.x compliant) and ESP8266 architectures.
- Lighthouse Auto-Discovery: Implements Zero-Touch Provisioning via SSID injection. Edge nodes actively scan for the Gateway's dynamic MAC address and Wi-Fi channel, bypassing static network configurations.
- Delay-Tolerant Networking (DTN): Utilizes LittleFS on the Gateway to buffer ESP-NOW payloads into local flash storage during upstream network outages, preventing data loss.
- Cloud-Native Observability: Routes telemetry via a Mosquitto MQTT broker into a Telegraf and InfluxDB pipeline, visualized in real-time through a Grafana Command Center hosted on a K3s cluster.
The Gateway connects to the upstream network and broadcasts a dynamic Lighthouse beacon. The Edge Node powers on, scans the environment, extracts the MAC address from the Lighthouse signature, and dynamically tunes its radio channel to initiate the connection.
Gateway Node initializing the Lighthouse beacon, establishing the MQTT connection, and receiving initial edge data.
Edge Node executing the channel sweep, locking onto the Gateway's dynamic MAC, and transmitting telemetry.
Data is transmitted asynchronously via the ESP-NOW protocol. The hardware provides instant delivery acknowledgments to verify the successful routing of SOS signals.
Edge Node logging successful delivery when the Gateway is active, and delivery failure when the Gateway is offline.
If the upstream Kubernetes server disconnects, the Gateway intercepts the failure and buffers incoming radio packets into onboard flash storage.
The upstream MQTT connection drops. The Gateway detects the failure and buffers incoming edge data into LittleFS.
The server is restored. The Gateway instantly re-establishes the connection and flushes the buffered payload sequence to the database.
The backend relies on a TIG stack (Telegraf, InfluxDB, Grafana) to provide a fault-tolerant, real-time map of hardware health and survivor telemetry.
The Grafana dashboard tracking continuous analog metrics (Battery/CPU) with shifting color gradients, currently displaying a stable 'NORMAL' operational state.
The dashboard instantly switching to a critical 'EMERGENCY!' state alert upon receiving an active SOS payload from a decentralized Edge Node.
This repository represents Version 1.0 of the V.O.I.D. infrastructure. The current release focuses strictly on establishing the routing logic, auto-discovery protocols, and DTN architecture.
Notes on v1.0:
- Simulated Hardware Metrics: For this initial deployment, the
batteryPctandcpuLoadmetrics are generated via a randomized software simulation loop to stress-test the data pipeline. - Future Hardware Integration: Version 2.0 will bind these variables to physical hardware sensors (e.g., analog voltage dividers for LiPo batteries).






