An autonomous line-following and object-delivery robot that simulates agricultural logistics, built for the e-Yantra Robotics Competition (eYRC 2025-26) conducted by IIT Bombay.
🏆 Achieved Level 2 — Certificate of Completion | Top 10–30 ranked team nationally
Bird's-eye view of the electronics layer — eYRC 2025-26 CropDrop Bot custom PCB,
STM32 Nucleo F103RB, buck module (voltage regulation), Pro:Range 2200mAh 3S LiPo
battery, toggle power switch, and electromagnet mounted on the laser-cut acrylic chassis
Fully assembled CropDrop Bot — dual-layer circular acrylic chassis (Fusion 360 design,
laser cut), N20 drive motors with rubber wheels, electromagnet at front, eYRC custom PCB
and STM32 Nucleo on the top layer
Front sensor array — TCS3200 color sensor (LED ring visible) for crate color
identification, IR sensor module for line detection, N20 motor with rubber wheel,
and the cylindrical electromagnet for crate pickup and drop
| Field | Details |
|---|---|
| Competition | e-Yantra Robotics Competition (eYRC) |
| Edition | 2025–26 |
| Theme | CropDrop Bot |
| Organized by | ERTS Lab, Dept. of Computer Science & Engineering, IIT Bombay |
| Institution | Amrita Vishwa Vidyapeetham, Amritapuri, Kerala |
| Achievement | Level 2 — Certificate of Completion |
| Rank | Top 10–30 teams nationally |
| Sponsored by | MHRD, Government of India (NMEICT) |
| Level | Certificate | Description |
|---|---|---|
| Level 1 | Certificate of Merit | Top 1–10 teams |
| Level 2 ✅ | Certificate of Completion | Top 10–30 teams — Our Achievement |
| Level 3 | Certificate of Participation | Partial completion |
| Level 4 | Letter of Participation | Basic participation |
CropDrop Bot simulates an autonomous agricultural logistics robot that:
- Starts at a designated supply depot
- Navigates a white-line track representing farm pathways (curves, zigzag sections, and dotted/discontinuous segments)
- Detects and collects color-coded crates (each with a specific delivery priority)
- Transitions to a black-line track simulating village transportation lanes
- Identifies the correct Drop Zone for each crate and delivers it precisely
- Reliable line tracking across varying path geometries (curves, zigzags, dotted lines)
- Accurate color-based crate identification using a TCS3200 color sensor
- Precise pickup and drop actuation using an electromagnet
- Efficient task sequencing with PID/RL-based control
PID Control · Reinforcement Learning · Python · Embedded-C · Build-a-Bot
| Member | Role & Contribution |
|---|---|
| Manjari M | Reinforcement Learning (RL) — trained the RL agent for navigation decisions |
| Tanishk Choudhary | Turning Logic — implemented reliable turning at intersections and corners |
| Viswajit Arunkumar | Mechanical Design + Box Color Identification + Pickup Logic — built the robot chassis, integrated TCS3200 color detection, and programmed electromagnet pickup/drop |
| Tharaneeswarar MS | PID Control for Line Following — tuned and implemented the PID controller for smooth and accurate line tracking |
┌──────────────────────────────────────────────────────────────────┐
│ CropDrop Bot System │
│ │
│ Start at Supply Depot │
│ ↓ │
│ White-Line Tracking (PID Controller) │
│ ├── Straight paths │
│ ├── Curves & Zigzags │
│ └── Dotted/Discontinuous segments │
│ ↓ │
│ Color Sensor (TCS3200) → Identify Crate Color/Priority │
│ ↓ │
│ Electromagnet → Pick Up Crate │
│ ↓ │
│ Black-Line Tracking → Navigate to Drop Zone │
│ (Turning Logic + RL-based decisions) │
│ ↓ │
│ Identify Correct Drop Zone → Release Crate │
│ ↓ │
│ Return to Depot / Next Task │
└──────────────────────────────────────────────────────────────────┘
| Component | Qty | Purpose |
|---|---|---|
| STM32 Nucleo F103RB | 1 | Main microcontroller |
| STM32 mini USB Cable | 1 | Programming & power |
| N20 Motor | 2 | Drive wheels |
| Motor Driver L298N | 1 | Motor speed & direction control |
| N20 Wheel | 2 | Drive wheels |
| Mounting Bracket Motors | 2 | Motor mounting |
| Castor Wheel | 1 | Front support wheel |
| Line Following Sensor | 1 | White/black line detection |
| TCS3200 Color Sensor | 1 | Crate color identification |
| IR Sensor Module | 1 | Obstacle / proximity detection |
| Electromagnet | 1 | Crate pickup and drop |
| Mosfet Driver | 1 | Electromagnet switching |
| LiPo 3-cell 12V Battery | 1 | Power supply |
| RGB LED | 1 | Status indicator |
| Battery Balance Charger | 1 | Battery management |
| Xt60 Male Wire Extender (15cm) | 1 | Battery connection |
| Battery Voltage Tester | 1 | Voltage monitoring |
| Jumper Wires (M-M, M-F, F-F) | 60 | Connections |
| CropDrop Bot PCB | 1 | Custom PCB for integration |
| Buck Module (on PCB) | 1 | Voltage regulation |
| Toggle Switch (on PCB) | 1 | Power on/off |
| Resistor 300Ω (on PCB) | 3 | Current limiting |
| Studs, Nut & Bolt | 6 studs / 20 screws | Chassis assembly |
| Foam Boxes | 6 | Crates for delivery task |
| A4 Chart Paper (Red, Green, Blue) | 2 | Color identification targets |
| 5×5 cm Metal Sheet | 6 | Electromagnet pickup targets |
The robot chassis was designed from scratch using Fusion 360 and fabricated as a dual-layer laser-cut acrylic structure. The two-tier layout separates the battery and motor layer (bottom) from the electronics and sensor layer (top), ensuring clean wiring, easy access for maintenance, and a stable low center of gravity during fast line-following maneuvers
The TCS3200 color sensor reads RGB frequency values from the crate surface:
- Compares readings against calibrated thresholds for Red, Green, and Blue crates
- Maps each color to its delivery priority
- Triggers the electromagnet (via MOSFET driver) to pick up the metal sheet on the crate
- Releases the electromagnet at the correct drop zone
Priority Weighted Sequence is given based on following formula: ---> Priority = Summation[Package Weightage] * Box Picked and Placed. Where the weightage is as follows: R:50 | B:30 | G:10. and the Priority sequence is the number of correct packages delivered in the continuous sequence of the entire theme; The Priority Sequence is R-B-G
Step 1 — Startup: Robot powers on at the supply depot. RGB LED indicates ready state.
Step 2 — Farm Path Navigation: PID/RL controller follows the white-line track through curves, zigzags, and dotted segments.
Step 3 — Crate Detection: At each crate location, the TCS3200 reads the crate color. Priority is determined (Red > Green > Blue or per task specification).
Step 4 — Pickup: Electromagnet is energized via MOSFET driver, attracting the metal sheet on the crate base and lifting it.
Step 5 — Village Lane Navigation: Robot transitions to the black-line track. RL agent and turning logic handle intersection decisions toward the correct Drop Zone.
Step 6 — Delivery: Robot identifies the correct Drop Zone by color/marker, de-energizes the electromagnet, and releases the crate.
Step 7 — Repeat: Returns for the next crate in priority order until all deliveries are complete.
📄 The official e-Yantra Certificate of Completion (Level 2) is available in this repository:
👉 certificate/eYRC_2025-26_CropDropBot_Certificate.pdf
Certificate Details:
- Recipient: Viswajit Arunkumar
- Institution: Amrita Vishwa Vidyapeetham, Amritapuri, Kerala
- Competition: e-Yantra Robotics Competition (eYRC 2025–26)
- Theme: CropDrop Bot
- Level: Level 2 — Certificate of Completion (Top 10–30 nationally)
- Issued by: Prof. Shivaram Kalyanakrishnan, Principal Investigator, e-Yantra, IIT Bombay
- Certificate ID: EYRC2025-261776172716.80626BQJBIDI
| Category | Technology |
|---|---|
| Microcontroller | STM32 Nucleo F103RB |
| Firmware Language | Embedded-C |
| High-level Logic | Python |
| Line Following | PID Control |
| Navigation Decisions | Reinforcement Learning |
| Color Detection | TCS3200 (RGB frequency sensing) |
| Actuation | Electromagnet + MOSFET Driver |
| Motor Control | L298N Motor Driver |
| Power | LiPo 3-cell 12V + Buck converter |
| Custom PCB | CropDrop Bot PCB |
Developed for e-Yantra Robotics Competition 2025–26 · IIT Bombay · Team: Manjari M, Tanishk Choudhary, Viswajit Arunkumar, Tharaneeswarar Ms