2. ఉప అంశాలు
-
AI for Better Life.
-
Green Energy & Environmental Conservation.
-
Mathematical Modelling and Computational Thinking.
-
Health, Hygiene and Sanitation.
-
Indigenous Agricultural Practices.
-
Science for Rural Transformation.
-
Disaster Management.
3. విభాగాలు
-
Junior Category: Classes VI, VII and VIII.
-
Senior Category: Classes IX to XII.
Project Breakdown (Nos. 46–50)
46. RoboAssist AI — Intelligent Human-Following Service Robot with Voice Interaction and Obstacle Avoidance
-
Core Capabilities: Tracks a target person smoothly, navigates around static and moving obstacles, and responds to spoken commands.
-
Suggested Stack: Jetson Orin Nano or Raspberry Pi 5 + ROS 2; RGB-D camera or depth sensor + ultrasonic/LiDAR array; person detection/re-ID model paired with local or cloud speech-to-text, LLM/intent routing, and text-to-speech.
-
Demo & Academic Angle: Show live transitions between follow mode, avoidance rerouting, and voice-triggered delivery/assistance, while benchmarking tracking latency and obstacle clearance distance.
47. MineSafe AI — AI-Based Mining Safety Monitoring and Hazard Detection Using IoT Sensors
-
Core Capabilities: Monitors underground or open-pit conditions in real time to detect toxic/flammable gas buildup, temperature spikes, structural vibration, or worker proximity hazards.
-
Suggested Stack: ESP32 / STM32 sensor nodes with LoRaWAN or Wi-Fi/MQTT; MQ-series/electrochemical gas sensors, temperature/humidity, IMU/vibration, and wearable/beacon tags; anomaly detection (Isolation Forest, Autoencoders, or LSTM) feeding a live alert dashboard.
-
Demo & Academic Angle: Simulate gas/heat threshold breaches and abnormal vibration patterns in a scale model or test rig, highlighting low-bandwidth telemetry and fail-safe local alarms when connectivity drops.
48. IndustryGuard 4.0 — Intelligent Industrial Machine Condition Monitoring and Predictive Maintenance Using Edge AI
-
Core Capabilities: Detects early bearing wear, imbalance, misalignment, or thermal drift on rotating machinery directly at the edge before unplanned downtime occurs.
-
Suggested Stack: MEMS accelerometers/vibration sensors + current/temperature sensors; Edge runtime (TensorRT, ONNX Runtime, or TFLite) on a Jetson/industrial gateway; FFT/spectral feature extraction with 1D-CNN, Autoencoder, or gradient-boosted models estimating remaining useful life (RUL).
-
Demo & Academic Angle: Induce controlled imbalance or load changes on a motor/fan testbed and compare raw time-series vs. frequency-domain edge inference accuracy and latency.
49. VisionSort AI — Computer Vision-Based Automatic Object Identification and Robotic Sorting System
-
Core Capabilities: Identifies parts by class, defect status, or color/shape on a conveyor or pick zone and directs a robotic arm or actuator to sort them into designated bins.
-
Suggested Stack: Overhead industrial/USB camera + conveyor/actuated sorting gate or multi-axis robotic arm (inverse kinematics / serial or CAN control); YOLOv8/YOLOv11 or custom CNN classifier with pixel-to-world coordinate calibration.
-
Demo & Academic Angle: Run a live mixed-item stream showing bounding-box confidence, misclassified/defective item rejection, and throughput (items per minute) across varying belt speeds.
50. PowerGuard AI — AI-Based Smart Electrical Fault Detection, Energy Monitoring and Predictive Protection System
-
Core Capabilities: Tracks voltage, current, power factor, and harmonics to spot short circuits, overloads, arc/earth faults, or abnormal load signatures and trigger protective cutoff.
-
Suggested Stack: Current/voltage transformers or smart metering ICs (e.g., ZMPT101B / SCT-013 / ADE-series) + microcontroller/edge node + relay/contactor trip circuit; waveform/harmonic feature extraction with anomaly/fault classifiers and load forecasting.
-
Demo & Academic Angle: Demonstrate real-time energy dashboards alongside safe low-voltage fault injection (overload, phase/line fault simulation) and fast automated trip response.
Comparison at a Glance
| Project | Primary Sensing | Core AI / Analytics | Main Output / Action |
| RoboAssist AI | Depth/RGB camera, LiDAR/ultrasonic, mic | Person tracking, path planning, voice AI | Wheel motion, obstacle avoidance, voice reply |
| MineSafe AI | Gas, temperature, vibration, proximity IoT | Time-series anomaly & hazard detection | Siren/beacon alerts, dashboard telemetry |
| IndustryGuard 4.0 | Vibration, current, temperature sensors | Spectral/edge ML, health/RUL scoring | Maintenance alerts, condition health index |
| VisionSort AI | Vision camera, encoder/limit sensors | Object detection & classification | Arm/actuator pick-and-place or bin diversion |
| PowerGuard AI | Voltage, current, power/harmonic sensors | Fault classification & load prediction | Automated trip control, energy analytics |


No comments:
Post a Comment