Certified by the International Journal of Multidisciplinary Research
Developed using Arduino, FPGA, IoT, and Deep Learning
A 25–28 kg tracked robot that autonomously detects, verifies, and extinguishes indoor fires.
It fuses data from flame, smoke, temperature/humidity, PIR, ultrasonic, and UWB radar sensors with infrared-camera frames.
A custom CNN deployed on an FPGA confirms real fires (rejecting false positives like hot objects or sunlight).
A second CNN for live-human detection was designed as a v2 concept but not yet implemented.
Prototype-1 achieved over 95% accuracy in lab fire-detection tests.
Figure 1: High-level block diagram of the firefighting robot.
- Master Arduino UNO aggregates data from all fire-detection sensors and issues commands
- Slave UNO #1 drives the track motors via Cytron MDD3A drivers
- Slave UNO #2 controls the 2-axis water-nozzle and 3-axis camera servos via a PCA9685 board
- FPGA performs CNN-based fire confirmation using infrared-camera frames
- ESP8266 #1 handles remote control & telemetry through the Blynk cloud
- ESP8266 #2 serves as a backup telemetry channel
- A 2 HP centrifugal pump is switched on by a relay under Master/Slave control
Figure 2: Front and side blueprint of the robot chassis (dimensions in cm).
-
AI-Based Fire Confirmation
- CNN trained on infrared-camera images + multi-sensor data
- FPGA-accelerated inference for real-time decision-making
-
Human-Detection Concept (v2)
- Separate CNN + computer-vision pipeline to detect humans and assess life signs (concept stage only)
-
Master-Slave Microcontroller Network
- 1 × Master Arduino UNO (sensor fusion & logic)
- 2 × Slave Arduino UNO (motors & servos)
- 1 × Arduino Mega (testing routines only)
-
IoT & Telemetry
- 2 × ESP8266 modules running Blynk for remote monitoring and live video
-
Mobility & Extinguishing
- 6-wheel caterpillar tracks with front shock absorbers (stair-climbing)
- Orange OG555 DC motors + 3 × Cytron MDD3A drivers
- 2 HP centrifugal pump + high-pressure nozzle (2-axis servo)
Unlisted demo of Prototype-1 in action
-
📰 Published Research Paper
Siddharth Singh, Achyut Pandey, Harsh Vardhan Nilaratna, Manik Gurjar.
“Designing and Prototyping of an Autonomous Fire Fighting Vehicle”
International Journal for Multidisciplinary Research (IJFMR), Vol. 6, Issue 6, Nov–Dec 2024.
👉 Download Full Paper (PDF) -
🏆 Certificate of Publication
- Embedded Systems (Arduino, ESP8266, FPGA)
- Real-Time Sensor Fusion & Control
- Deep Learning & Computer Vision (CNN for fire confirmation; human-detection concept)
- IoT Integration (Blynk App)
- Autonomous Navigation & Fire Suppression
firefighting-robot-public/
├── media/
│ ├── architecture.png # Block diagram image
│ ├── robot_front.jpg # Prototype photo
│ └── demo-video.mp4 # Unlisted YouTube demo
│
├── Certificates/
│ ├── Siddharth_Singh_IJFMR_Certificate.jpg
│ └── Autonomous_Firefighting_Robot_Siddharth_Singh_IJFMR2024.pdf
│
├── Overview/
│ ├── Project_Summary.md # Problem statement, objectives, impact
│ └── Key_Features.md # High-level bullet points
│
└── README.md # This file
---
## 👥 Contributors
This project was developed by a four-member team as part of our final year engineering project.
- **Siddharth Singh** – Embedded Systems, AI Integration (FPGA & Sensors), Testing, and GitHub Repository
- Achyut Pandey – Mechanical Design, Frame & Mobility System
- Harsh Vardhan Nilaratna – Documentation, Research Paper, Safety Analysis
- Manik Gurjar – IoT Integration (Blynk), Sensor Calibration, Troubleshooting
> 📌 The GitHub repository was created and maintained by **Siddharth Singh** to publicly showcase the team’s work.
---
> ⚠️ **Note**: Source code, detailed schematics and raw data are kept private to protect intellectual property. Recruiters may request access on demand.