TTKLabs / Projects / Atlas
Hardware + AI // Environmental Monitoring

PROJECT
ATLAS

Status
Development — Prototyping
Started
Jan 2026
Category
IoT
Scale
City-Wide Deployment
// Overview

An open-source environmental monitoring platform that combines custom sensor hardware with real-time AI analysis — tracking air, water, and soil health at city scale.

Atlas started from a simple question: why is it so hard to know what’s actually in the air you breathe? Existing monitoring infrastructure is expensive, sparse, and often locked behind proprietary systems. We set out to change that.

The platform consists of weatherproof sensor nodes built around custom PCBs, a low-power mesh network for data transmission, and a cloud-connected AI layer that interprets readings in real time — flagging anomalies, predicting trends, and surfacing actionable insights for city planners and residents alike.

// By The Numbers

IMPACT AT SCALE.

1
Sensor Nodes Deployed
12
Environmental Metrics
99.2%
Uptime Over 90 Days
3s
Data Latency
// Core Features

WHAT ATLAS DOES.

01
Multi-Sensor Nodes
Each node monitors PM2.5, PM10, CO‍₂, NO‍₂, temperature, humidity, UV index, and soil moisture. Modular sensor slots allow field upgrades without replacing the base unit.
02
Mesh Network
Nodes form a self-healing LoRaWAN mesh. If one node goes offline, its neighbours reroute data automatically — ensuring continuous coverage across the deployment zone.
03
AI Anomaly Detection
A lightweight transformer model runs continuous analysis on incoming streams. It distinguishes pollution spikes from sensor noise, correlates multi-location events, and issues real-time alerts.
04
Public Dashboard
Live maps, historical trends, and downloadable datasets — open to the public. Built for journalists, researchers, and curious residents equally.
05
Solar + Battery Power
Each node is fully off-grid. A 6W solar panel charges a 10,000mAh LiPo pack, giving 72+ hours of autonomous runtime during low-light periods.
// Technology Stack

BUILT WITH.

Hardware
Custom PCB
4-layer board, STM32 MCU, onboard power management
Firmware
Embedded C / Zephyr RTOS
Low-power scheduling, OTA update support
Connectivity
LoRaWAN + MQTT
TTN gateway integration, TLS-encrypted uplink
AI / ML
PyTorch + ONNX
Time-series transformer, edge-deployable inference
Backend
TimescaleDB + FastAPI
Time-series ingestion, REST + WebSocket APIs
Frontend
SvelteKit + Mapbox
Real-time map tiles, responsive public dashboard

CONTRIBUTE OR
DEPLOY ATLAS.

Everything is open source. Fork it, deploy it, improve it.

View on GitHub →Get In Touch
Next Project
PULSEKIT
→