This repository documents the design and development of a low-cost, feature-rich AMR (Autonomous Mobile Robot) platform intended for indoor environments such as warehouses, factories, and hospitals. The robot combines real-time sensing, robust motor control, modular construction, and remote-monitoring capabilities—engineered by undergraduates at the Department of Electronic & Telecommunication Engineering, University of Moratuwa.

This project was undertaken for EN2160 – Electronic Design Realization as a collaborative team assignment. The aim was to replicate—then innovate on—the commercial Omron LD-60 AMR while keeping costs low.

✨ Key Objective: Build a reliable mobile robot capable of autonomous material handling in structured indoor environments.

• ✅ Autonomous Navigation with obstacle detection (LakiBeam 1 S LiDAR)
• ✅ Jetson Nano on-board SLAM & path planning
• ✅ Bare-metal Atmega32U4 firmware for deterministic timing and high-precision motor control
• ✅ Real-time Telemetry over UDP with future fleet-management scalability
• ✅ Remote Controlling for initial Mapping of the environment over WiFi
• ✅ Closed-loop NEMA 24 Steppers with for precise, high-torque motion
• ✅ 7.0' USB Touch Display providing a user-friendly on-robot interface
• ✅ Efficient DC-DC Power Conversion for motor drivers and logic systems
• ✅ Modular Aluminium & Steel Enclosure with an industrial aesthetic

Key parameters

• Dimensions : 82.5 cm × 60 cm × 20 cm (L × W × H)
• Gear Ratio : 10 : 1
• Wheel Diameter : 20 cm
• Wheel Configuration : 2-wheel differential drive + 4 castors

View	Description
	Isometric View from front
	Back view
	Isometric view from back
	Top view
	Side View
	Wheel layout
	Wheel & motor configuration
	Steel chassis

The CAD model is for measurement verification & demonstration only. Not direct manufacturing. The coupler, Back motor Mount, and enclosing lid's designs were altered in the manufactured assembly.

Component	Preview
Custom Motor Coupler
Steel Chassis
Drive Wheel
Modified Castor
Motor Placement

Acquiring encoder readings from closed-loop stepper motors for sensor fusion and LIDAR-based mapping presented significant challenges. The quadrature encoder signals are typically designed to interface directly with the motor controller, making MCU integration complex.

Method: Split encoder signal lines using Y-junctions, directing one branch to the motor controller and another to the MCU.

Results:

• Encoder counts showed significant offset from actual readings
• Motors exhibited vibration and jerky movement
• Signal integrity compromised due to impedance mismatch

Method: Implemented 4-channel single-mode buffering using SN74HC125N ICs for all encoder channels (A+, A-, B+, B-).

Quadrature encoder signal from 2 channels	Before and after buffering, Amplitude is reduced at the output unexpectedly

Specifications:

• Buffer selected to match encoder slew rate and voltage levels
• Designed for proper fan-out capability
• Target frequency: Up to 3600 RPM motor operation

Results:

• Performance remained unsatisfactory
• Issues likely attributed to low-quality, locally sourced buffer ICs

Method: Combined imported SN74HC125N buffers with SN75157DR differential receiver ICs.

Results:

• ✅ Accurate encoder reading up to 500 RPM with zero count loss
• ⚠️ Above 500 RPM: ~3 count loss per 2000 counts
• ❌ Count loss increased approximately quadratically with every 100 RPM increment

Initial 15cm diameter wheels provided insufficient ground clearance when integrated with the motor and chassis assembly, limiting the robot's ability to traverse obstacles.

• Wheel Upgrade: Increased diameter from 15cm to 20cm
• Custom Coupling: Designed and manufactured a custom motor coupler
• Ground Clearance: Achieved >4cm clearance to meet traversal requirements

Problem: Rapid development cycle led to suboptimal pin assignments, resulting in one motor being controlled by a 10-bit timer lacking hardware-controlled CTC (Clear Timer on Compare) mode.

Initial Workaround:

• Generated a 50 % duty-cycle “pseudo-step” PWM. To span the full RPM range, we had to switch prescalers on-the-fly - acceptable for the first milestone, but sub-optimal.

Solution:

• Dual prescaler configuration to cover entire RPM range
• Successfully met first evaluation requirements

• Encoder Pin Mapping: Incorrect pin assignments for encoder interfaces
• USB Communication: Instability due to missing external oscillator for MCU timing

PCB Revision: A new PCB revision is planned to address:

• Corrected pin mapping for all motor and encoder interfaces
• Integration of external oscillator for stable USB communication
• Implementation of lessons learned from current design iteration

These challenges provided valuable insights into signal integrity, mechanical design constraints, and the importance of careful PCB layout planning in robotics applications.

• Implement Custom SLAM and path planning alogorithem
• Implement velocity profiling in for the lower layer controller
• Develop an remote debugging and controlling interface for the robot

Board	Preview (Top / 3D / Assembled)	Summary
Microcontroller PCB (REV 1)		ATmega32U4–based board that: • samples IMU data, battery voltage, and cliff-sensor inputs; • drives stepper motors via pulse-direction outputs; • converts differential encoder signals to single-ended with on-board line receivers; • exposes a USB-C port (USB-CDC) to the Jetson Nano; • provides an interrupt line for the emergency-stop button.
Power Distribution PCB (REV 1)		Steps the 22 V battery rail down to 48 V, 12 V, 5 V, and 3.3 V using plug-in buck-converter modules and routes the main supply through the on-board power switch.
Encoder Buffer PCB		Buffers all eight quadrature-encoder channels (two motors) and actively drives them so the signals can be split to both the MCU and the stepper-motor drivers without loading or skew.
IMU Adapter PCB		Provides modular mounting options for the IMU, to minimise magnetic interference and easy replacement; connects to the main board via a dedicated ISP adapter cable.

• Estimated Total Weight : 28.8 kg
• Output Torque @ Wheel : ≈ 6.0 N·m
• Acceleration : ≈ 5.88 m s⁻²
• Displacement Accuracy : ≈ 2.4 mm / step (Changed)

• 📘 Current Design Report (PDF)
• 📦 Image Assets

• 🔋 Battery Management System (BMS) with SoC tracking
• 📦 Swappable payload modules
• 🧭 Autonomous docking & charging
• 🌐 Full ROS 2 support with fleet-level coordination
• 📊 Cloud dashboard & OTA updates

• 🏭 Warehouse logistics automation
• 🏥 Hospital supply transport
• 🛍️ Retail floor delivery
• 🎓 Academic research labs

This prototype balances industrial performance with academic accessibility, demonstrating that carefully-chosen low-cost components can rival premium AMR solutions.

Designed by students, built for the future of autonomous mobility.