A compact, low-cost Autonomous Mobile Robot (AMR) built for real-world MSME factory deployment
This repository is published strictly for portfolio, demonstration, and evaluation purposes. No source code, firmware, wiring schematics, or design files are shared publicly. Reproduction, redistribution, reverse-engineering, or commercial/academic reuse of this project — in whole or in part — is not permitted without explicit written consent from the team. This repository exists to showcase the system, not to enable rebuilding it.
MechMover is a task-specific Autonomous Mobile Robot designed to solve a real, unglamorous factory problem: moving 10–20 kg loads through narrow indoor aisles, safely, without the cost of LiDAR-based commercial AMRs.
Manual material handling is slow and unsafe. Fixed-path AGVs can't adapt to obstacles. Commercial AMRs are priced out of reach for small and medium manufacturers. MechMover was built to close that gap — with intelligent local navigation instead of expensive global mapping, and mechanical design robust enough for daily industrial use.
🏆 1st Prize — Pragyan 2026, NIT Trichy (Industrial Automation Theme, SAE India Guided)
MechMover physical prototype — six-wheel chassis with dual-layer sensor suite
- Manual material handling is time-consuming and unsafe
- Narrow factory aisles restrict the use of large robots
- Existing AGVs follow fixed paths and cannot adapt to obstacles
- Commercial AMRs using LiDAR are too expensive for MSMEs
Need: A compact, safe, and low-cost autonomous material handling solution.
- Design a low-cost autonomous mobile robot
- Navigate narrow indoor factory aisles
- Operate safely around humans and machinery
- Transport 10–20 kg payloads reliably
- Implement obstacle avoidance and basic path planning
- Prioritize stability, simplicity, and reliability over raw complexity
MechMover follows a Brain–Body separation architecture across four independently optimized layers:
| Layer | Role |
|---|---|
| Mechanical Platform | Six-wheel, center-drive chassis with low center of gravity |
| Perception Layer | Proximity and IR-based sensing for obstacle detection & docking |
| Control & Intelligence Layer | Raspberry Pi (decision-making) + Arduino Mega (real-time motor control) |
| Power & Energy Layer | Single high-capacity battery with regulated supply for electronics |
| Raspberry Pi — Brain | Arduino Mega — Body |
|---|---|
| High-level decision making | Real-time motor control |
| Processes navigation logic | Reads sensors continuously |
| Selects movement direction | Executes motion commands |
| Interfaces with fleet manager | Generates precise PWM signals |
- 🔄 Real-time adaptive obstacle avoidance (no fixed paths)
- 💰 Low-cost sensor system — no LiDAR required
- 🧭 Memory-based path optimization (avoids previously blocked routes)
- 📦 20 kg payload capacity with uniform load distribution
- 🎯 Precision IR-based docking, accurate to ±5 cm
- 🏭 Fleet-ready architecture — scales from single unit to multi-robot deployment
- 🔧 Compact footprint — no factory layout modification required
- Continuously sense the environment (front/left/right clearance)
- If front distance ≥ 30 cm → move forward
- If front distance < 30 cm → compare left/right clearance and select the safer path
- Avoid previously blocked directions (adaptive memory)
- On approaching a target station, IR docking markers guide precision alignment
- Repeat continuously for autonomous, unsupervised operation
Implementation details, sensor calibration values, and control firmware are intentionally not included in this repository.
- Rigid chassis suitable for continuous factory-floor operation
- Six-wheel configuration for stability under load
- Center-drive propulsion mechanism
- Flat payload deck supporting bins, trays, and small pallets
- Design optimized for safe acceleration and braking, not top speed
- Centralized task allocation across multiple robots
- Multi-robot coordination and congestion avoidance
- Status monitoring and operational logging
- Designed for scalability: single robot → fleet deployment
- Inter-workstation material transfer
- Assembly line logistics
- In-plant transport for MSMEs
- Educational and research lab environments
(Video demo added below)
| Requirement | Status |
|---|---|
| Compact design | ✅ |
| Low cost | ✅ |
| Indoor navigation | ✅ |
| 10–20 kg payload | ✅ |
| Basic path planning | ✅ |
| Docking capability | ✅ |
| Simplicity & reliability | ✅ |
- Vision-based navigation
- Advanced fleet analytics dashboard
- Autonomous charging docks
- Integration with MES/ERP systems
| Name | Department |
|---|---|
| Sugavanam M | Electronics & Communication Engineering |
| Nishanth M | Electronics & Communication Engineering |
| Siva Santhosh V | Electronics & Communication Engineering |
| Naren V T | Electronics & Communication Engineering |
Department of Mechanical Engineering, Mahendra Institute of Technology, Namakkal
This project is released under a restricted, all-rights-reserved license. No part of this repository — including design concepts, documentation structure, or visual assets — may be copied, forked for reuse, or repurposed for another project or submission without prior written permission from the team.
See LICENSE for full terms.
MechMover — proving that industrial automation doesn't need expensive sensors, just intelligent design decisions.
Pragyan 2026 · MIT Namakkal · MechMover Team