SIH26039HardwareSmart Automation
₹1,50,000 Prize (₹75k Joint)
•
Deadline: 20 September 2026
•
Submitted Ideas: 0/500

Al-Powered Underground Mine Safety, Monitoring and Rescue System.

Ministry / Organization: Governmcnt of Jharkhand
Department: Department of Higher & Technical Education
Technical Complexity
8/10
Dataset Accessibility
7/10
36h Hackathon Viability
8/10
Step 0 • Zero-Jargon Layman Summary

“Underground coal mines in Jharkhand face severe hazards like toxic gas leaks, cave-ins, and darkness, putting rescue workers at high risk. We need to build an intelligent robotic rover or drone that can autonomously explore dangerous tunnels, detect toxic fumes and structural threats, spot trapped miners using thermal cameras, and beam live data back to a safe surface control station.”

The DecodeX PS Decoder™

Deconstructed Problem Blueprint

Step 1

WHAT TO BUILD?

A rugged, remote-controlled and semi-autonomous robotic ground rover (or compact drone) paired with a surface-level web command-and-control dashboard running real-time AI analytics for hazard detection and victim tracking.

Step 2

WHO IS IT FOR?

Primary beneficiaries are underground coal miners and disaster response teams (such as DGMS, CMPFO, and NDRF). The users of the software and hardware system will be mine safety officers and rescue mission commanders at the surface control station.

Step 3

WHY DOES IT EXIST?

Current rescue methods rely on human rescuers entering unstable, toxic, and low-visibility underground environments blind, leading to delayed response times and catastrophic loss of human life among rescue personnel themselves.

Step 4

INPUT → PROCESS → OUTPUT ARCHITECTURE PIPELINE

01 • Inputs & Telemetry

Raw gas concentration readings (MQ-series/NDIR sensors for CH4, CO, O2), temperature & humidity values, LiDAR / depth camera point clouds, thermal and RGB optical video feeds, and IMU/wheel encoder telemetry.

02 • Processing & AI Core

Edge processing using ROS2 (Robot Operating System) for SLAM (Simultaneous Localization and Mapping), onboard YOLOv8 for human/survivor and obstacle detection, multi-sensor data fusion for gas anomaly tracking, and mesh-network signal relay back to the surface.

03 • Actionable Deliverables

A live surface dashboard displaying a real-time 3D point-cloud map of the mine, heatmaps of toxic gas leaks, automated hazard alert popups, and bounding-box telemetry pinpointing trapped workers with thermal signatures.

Step 5

CONSTRAINTS

  • Zero Wi-Fi/GPS connectivity underground, necessitating ruggedized node-to-node wireless mesh communication or tethered lines.
  • Explosive atmospheres requiring intrinsically safe electronics (ATEX/IECEx certified or purged enclosures to prevent spark ignition of methane gas).
  • Low visibility, extreme dust, and complete darkness requiring thermal imaging and active IR illumination.
  • Limited onboard battery life requiring optimized power management and low-power edge compute (e.g., NVIDIA Jetson Nano/Orin).
Step 6

JURY METRICS

  • Gas leak detection response latency (< 2 seconds from sensor trigger to UI alert).
  • SLAM mapping accuracy and drift rate in GPS-denied tunnels.
  • Object detection precision/recall for human/survivor localization in thermal video streams.
  • Robustness and packet-loss resilience of the wireless mesh transmission over a distance of at least 500 meters underground.
Step 7

CORE BOTTLENECK

Achieving reliable wireless data transmission and real-time video streaming through hundreds of meters of solid rock, coal pillars, and winding tunnels without traditional Wi-Fi or cellular infrastructure.

Why You Should Choose This PS (Pros)

  • Extremely high social impact addressing a real, life-and-death problem statement directly posed by the Government of Jharkhand.
  • Hardware theme with a physical prototype allows for high 'wow factor' during offline grand finale evaluations compared to pure web apps.
  • Clear, modular scope where software (dashboard + AI) and hardware (rover + sensors) can be divided efficiently among team members.

Challenges & Risks to Watch (Cons)

  • Hardware building carries physical risks of component failure, wiring errors, or damage during transit to the hackathon venue.
  • Simulating an underground environment (dust, toxic gases, zero GPS) inside a convention center hall requires creative testing hacks (e.g., smoke machines, gas lighters for methane simulation, dark tunnels built from cardboard).
  • Strict safety regulations around lithium batteries and electrical sparks during live demonstrations.
Hackathon Blueprint

Recommended Tech Stack & Grand Finale Features

Frontend & Mobile
Next.jsTailwindCSSThree.js / React Three Fiber (for 3D mine map visualization)
Backend & APIs
FastAPI (Python)WebSockets (for real-time telemetry streaming)MQTT broker (Mosquitto for IoT sensor ingestion)
AI / ML & Models
YOLOv8 / YOLOv11 (trained for thermal/RGB person detection)OpenCV (for image processing and dust filtering)Cartographer or RTAB-Map (for ROS2-based LiDAR SLAM)
Database & Cloud
PostgreSQL + PostGIS (for telemetry and spatial logging)InfluxDB (optimized for time-series gas and environmental sensor logs)

Killer Features to Impress the SIH Jury

Auto-Generated Gas Heatmap Overlay: Instantly interpolates sensor data across the 3D SLAM map to render a color-coded toxic plume visualization for rescue teams.
Autonomous Return-to-Home (RTH) Mesh Relay: If wireless signal drops below threshold, the rover autonomously backtracks along its mapped trajectory until signal is re-established.
Acoustic SOS Detection: Onboard microphone filtering background machinery hums to detect human vocal cries for help or whistle frequencies using lightweight FFT algorithms.
View Official SIH 2026 Portal Raw Text
▼
BACKGROUND

Jharkhand's underground coal mines face significant safety challenges, including toxic gas leaks,tunnel collapses, flooding, and poor visibility. During emergencies, rescue teams often lack real-time information about underground conditions, increasing risks and delaying response efforts.An intelligent robotic system capable of monitoring mine conditions, detecting hazards, and locating trapped workers can significantly improve mine safety and rescue operations while reducing risks to human rescuers.

DESCRIPTION

The Al-Powered Mine Safety and Rescue Rover is an inteltigent robotic system designed to operate in hazardous underground mining environments. The rover is equipped with gas sensors, thermal and night-vision cameras, environmental monitoring sensors, and wireless communication modules to provide real-time information about mine conditions.The system can detect toxic gases, monitor temperature and humidity, identify potential hazards, and assist in locating trapped workers during emergencies. Using Al-based analysis and remote monitoring capabilities, the rover enables rescue teams to assess underground conditions without exposing personnel to dangerous environments. The solution aims to enhance mine safety, improve emergency response efficiency, and reduce the risk of casualties in underground mining operations.

EXPECTED SOLUTION

Develop an Al-powered mine rescue system consisting of a rugged ground rover or a compact aerial drone capable of operating in hazardous underground mining environments. The system should provide real-time monitoring of toxic gases, temperature, humidity, and structural conditions while transmitting live video and thermal imaging data to a surface control station.

The rover and drone will assist rescue teams by exploring inaccessible areas, detecting hazards, locating trapped workers, and providing situational awareness during emergencies. The solution should improve mine safety, reduce risks to rescue personnel, and enable faster, more informed emergency response in underground mines.