Project-ARIA

Autonomous Robotic Integration for Automation

Introduction

Inefficiencies and safety risks in warehouses, healthcare, public spaces, manufacturing, infrastructure, and agriculture arise from reliance on manual and semi-automated systems. These systems lead to higher labor costs, operational delays, and errors. Advanced autonomous navigation solutions are needed to streamline operations, enhance accuracy, ensure safety, and improve scalability. Implementing such technology can significantly reduce operational costs, enhance the quality of service, and boost productivity across these sectors.

Overview

Our solution is a ROS-based autonomous navigation robot designed for versatile applications across various industries. It utilizes advanced sensors and algorithms to navigate indoor environments such as warehouses, hospitals, public spaces, manufacturing facilities, and agricultural sites. By automating routine tasks, our robot enhances operational efficiency, reduces human error, and improves safety. Its innovative design allows for seamless integration into existing workflows, scalable deployment, and adaptability to different use cases, making it a comprehensive solution for improving productivity and service quality across multiple sectors.

Table of Contents

• Demo
• Components
• Hardware
• Code Base
• Technologies Used
• Folder Structure
• Running A.R.I.A
• Result
• Conclusions

Demo

Demo Videos

Simulation Demo

sim_vid.mp4

Simulation Demo

Physical Demo

phy_vid.mp4

Physical Demo

rt_laser_scan.mp4

Real Time Lidar Data in GUI

rt_map.mp4

Real Time Mapping

rt_nav.mp4

Real Time Autonomous Navigation

Progress Images

Simulation

Fig: Image of ARIA in Simulation

Fig: Image of custom warehouse

Fig: Map of warehouse generated by SLAM Toolkit

Physical Prototype

Fig: Initial Prototype

Fig: Custom PCB for Nano & Motor Driver

Fig: ARIA Version 2 Chassis Base

Fig: Physical PCB

Fig: ARIA Version 2 (Without Top Cover)

Fig: Map generated in Real Environment

Fig: Cost Map for Autonomous Navigation

Fig: Marker to set Waypoint for ARIA To Navigate to

Fig: Graph of all the communication taking place between PI, dev Machine, different sensors & Controller

Components required with Bill of Materials

Components That are Already Acquired / Owned :

Component	Quantity	Description	Links to Products
Raspberry Pi 3b+	1	Microprocessor Board	Raspberry Pi 3b+
Pi cam	1	Standard Pi Cam for Surveillance	Pi camera
Buck Converter	1	Step down or Step up Voltage device	Buck Converter
Li-Po battery	1	11.5V Li-Po battery for Power Requirements	LiPo battery 3s
Lidar	1	RP Lidar A1M8 For Mapping the required area	Lidar A1M8
Caster wheel	1	For Smooth running and weight Balancing of the Bot	Castor wheel
Motor Driver	1	L298N Motor Driver to Run the Motors	Motor Driver
Arduino Nano	1	Microcontroller board for Sending the signal from Pi to Motor driver	Arduino Nano
Slide Switches	2	Switches for Powering Up or Down the Bot	Slide Switch
Wheels	2	To Run the Bot Forward, Reverse or in Required Direction	Wheels
Gear Motor Encoder D type	2	Gear motor Encoder D type to maintain or equalize the Speed of Motors	Encoded Motors

Hardware

Pinout Diagram

Table for Pin Connections

Component	Pin Configuration	Description
RP Lidar A1M8 Micro USB Port	USB Port 2	Lidar Connected to the Raspberry Pi’s USB Port 2
Arduino Nano Micro USB Port	USB Port 1	Nano Connected to the Raspberry Pi’s USB Port 1
Pi camera	CSI Camera Port	Camera Connected to the Raspberry Pi’s Camera port
Motor Driver control pin 1	PD6	CP1 of Motor driver connection to Microcontroller Nano
Motor Driver control pin 2	PD10	CP2 of Motor driver connection to Microcontroller Nano
Motor Driver control pin 3	PD9	CP3 of Motor driver connection to Microcontroller Nano
Motor Driver control pin 4	PD5	CP4 of Motor driver connection to Microcontroller Nano
Motor Driver 5V out	Vin	Power supply to Arduino Nano from Motor Driver
Encoder Motor1 +ve	OUT1	Motor1 Connections to Driver (To run the motor)
Encoder Motor1 -ve	OUT2	Motor1 Connections to Driver
Encoder Motor1 VCC	5v out	Power supply for E.motor from MotorDriver
M1 Encoder A (interrupt)	PC4	Connections to pin A4 of Microcontroller to determine the Speed of motor
M1 Encoder B (direction)	PC5	Connections to pin A5 of Microcontroller to determine the Direction and position of Motor
Encoder Motor2 +ve	OUT3	Motor2 Connections to Driver
Encoder Motor2 -ve	OUT4	Motor2 Connections to Driver
Encoder Motor2 VCC	5v out	Power supply for E.motor from MotorDriver
M2 Encoder A (interrupt)	PD2	Connections to pin D2 of Microcontroller to determine the Speed of motor
M2 Encoder B (direction)	PD3	Connections to pin D3 of Microcontroller to determine the Direction and position of Motor
GND pins of Encoder Motors	GND	Connections to GND pin of Motor driver
Buck converter out +	5v Vin	Reduced voltage power supply from BC to Pi
Li-Po Battery +ve	12v Vin of MD, In+ of BC	Supplying equal power to both Motor Driver and BC by making it short
GND Pins	Common Ground	All the GND Pins are shorted to make One common Ground

Code Base

Simulation

View Simulation Code

ROS Arduino Bridge

View ROS Arduino Bridge Code

Robot Code

View Robot Code

Technologies Used

• ros_arduino_bridge: Turns an Arduino into a motor controller! It provides a simple serial interface to communicate with a high-level computer running ROS, and generates the appropriate PWM signals for a motor driver, to drive two motors.
• ROS 2 Humble: The Robot Operating System (ROS) 2 Humble Hawksbill, a flexible framework for writing robot software.
• Gazebo: A powerful robot simulation tool that allows for testing and development in a simulated environment.
• SLAM Toolkit: Simultaneous Localization and Mapping (SLAM) toolkit used for mapping and navigation.
• Nav 2: The Navigation 2 stack in ROS 2, used for autonomous navigation of robots.
• OpenCV: Open Source Computer Vision Library, used for image processing and computer vision tasks.
• Rviz2: A 3D visualization tool for ROS 2, used to visualize the state of the robot and its environment.

The assests used for the simulation world (racks , boxes etc) is from an open source platform called AWS Robomaker . The world is a custom , developed by us .

View AWS Robomaker Repo

Folder Structure

Dev Machine Structure

dev_ws/
├── build/
├── install/
├── log/
└── src/
    ├── articubot_one/
    ├── serial_demo_msgs/
    └── ball_tracker/

Bot Workspace Structure

bot_ws/
├── build/
├── install/
├── log/
└── src/
    ├── articubot_one/
    ├── serial/
    └── different_drive_arduino/

Running A.R.I.A

Source the Workspace:

source install/setup.bash

Simulation

Run the following on the dev machine:

ros2 launch articubot_one launch_sim.launch.py world:=src/articubot_one/worlds/obstacles.world 

Physical Bot

To run the bot :

Run the following on the bot terminal:

ros2 launch articubot_one launch_robot.launch.py 

Run the following on the dev machine:

ros2 launch articubot_one joystick.launch.py 

To run the camera :

Run the following on the bot terminal:

ros2 launch articubot_one camera.launch.py

To run the lidar :

Run the following on the bot terminal:

ros2 launch articubot_one rplidar.launch.py

ros2 run rplidar_ros rplidar_composition --ros-args -p serial_port:=/dev/serial/by-id/usb-Silicon_Labs_CP2102_USB_to_UART_Bridge_Controller_0001-if00-port0 -p serial_baudrate:=115200 -p frame_id:=laser_frame -p angle_compensate:=true -p scan_mode:=Standard

To run Autonomous features :

GUI in Rviz2:

ros2 run rviz2 rviz2 -d src/articubot_one/config/main.rviz 

SLAM Map:

ros2 launch slam_toolbox online_async_launch.py params_file:=./src/articubot_one/config/mapper_params_online_async.yaml

For Autonomous navigation :

ros2 launch nav2_bringup navigation_launch.py params_file:=src/articubot_one/config/nav2_params.yaml

For Ball Tracking :

For Launching Camera :

ros2 launch articubot_one camera.launch.py 

For fixing compressing and decompressing vid :

ros2 run image_transport republish compressed raw --ros-args -r in/compressed:=/camera/image_raw/compressed -r out:=/camera/image_raw/uncompressed

For viewing feed :

ros2 run rqt_image_view rqt_image_view

For ball tracker :

ros2 launch ball_tracker ball_tracker.launch.py params_file:=src/articubot_one/config/ball_tracker_params_robot.yaml image_topic:=/camera/image_raw/uncompressed

Result

Our Project ARIA achieved significant milestones in autonomous navigation and operational efficiency. The robot successfully navigated through various indoor environments, including warehouses and simulated spaces, using advanced sensors and algorithms. Key accomplishments include:

• Accurate Mapping: Utilized SLAM Toolkit to generate precise maps of the environment, aiding in efficient navigation.
• Reliable Navigation: Implemented Nav2 stack for autonomous navigation, allowing the robot to follow designated paths and avoid obstacles.
• Effective Sensor Integration: Seamlessly integrated RP Lidar, Pi camera, and other sensors for real-time data collection and processing.
• Stable Communication: Ensured robust communication between the Raspberry Pi, Arduino Nano, and motor drivers through the ROS 2 framework.
• Simulation Success: Verified robot behavior and navigation algorithms in a simulated environment using Gazebo, ensuring readiness for real-world deployment.
• Operational Efficiency: Demonstrated the ability to automate routine tasks, enhancing productivity and reducing human intervention.

Conclusion

Project ARIA presents a versatile and scalable solution for autonomous navigation across various industries. By leveraging ROS 2 Humble, Gazebo, SLAM Toolkit, Nav2, OpenCV, and Rviz2, we developed a robust system capable of enhancing operational efficiency, reducing errors, and ensuring safety. The successful integration of advanced sensors and algorithms positions ARIA as a comprehensive tool for improving productivity and service quality in warehouses, healthcare, public spaces, manufacturing, and agriculture.

The project's success highlights the potential of autonomous robots in transforming traditional workflows, providing a strong foundation for future advancements and applications in autonomous technology.