🛒 UMM (Useful Market Mobility) – Autonomous Shopping Assistant for the Visually Impaired

An autonomous mobile robot designed to support independent shopping for visually impaired individuals.
By combining SLAM, voice recognition, and object detection technologies, this mobility solution enhances accessibility and autonomy.

ff

📌 Project Overview

• Project Title: Autonomous Mobility for Visually Impaired Shoppers
• Team Name: EMOM
• Department: Mechanical Engineering, Dankook University
• Team Members: Byeongchun Park, Sang Yoon, Hansol Jang, Joonmo Han
• Event: 2024 UMM Campus Capstone Design Competition

---

🔧 Hardware Components

• 🔹 2D LiDAR
• 🔹 Mecanum wheels + Encoder motors (PD control)
• 🔹 Arduino Uno & Mega
• 🔹 Jetson Nano
• 🔹 Camera (YOLOv5 object detection)
• 🔹 RFID Reader
• 🔹 Bluetooth module (voice control via smartphone)

---

🧠 Software Stack

• 🐍 Python (Jetson-side)
• ⚙️ Arduino (C) for motor & sensor control
• 🤖 ROS Melodic
• 🧭 SLAM with Cartographer + LaserScanMatcher
• 📷 YOLOv5 for object recognition
• 📐 AutoCAD & Fusion 360 for hardware design

---

🗺️ How It Works

1. Voice-Based Shopping Request

   • The user speaks the desired item into a smartphone app connected via Bluetooth.

2. Localization & Navigation

   • Cartographer SLAM is used to build a map.
   • Real-time position is estimated using LiDAR + Encoder Odometry.
   • move_base navigates to the item's location.

3. Item Identification

   • At the destination, the robot uses YOLOv5 to scan and detect the product via camera.
   • RFID is scanned for detailed product info.

4. Information Delivery

   • The item’s name and info are sent to the smartphone and spoken aloud via Bluetooth audio.

5. Return to Checkout

   • After scanning, the robot autonomously returns to the cashier area.

---

🧪 Key Innovations

• SLAM-based Navigation: Obstacle-aware indoor mapping and localization.
• PD Motor Control: Smooth omnidirectional mobility using Mecanum wheels.
• Voice-Driven UX: Full flow from voice command to voice feedback.
• YOLOv5 + RFID Fusion: Robust and accurate product identification.

---

🎥 Demo Video

---

📄 Documentation

For a detailed explanation of this project, please refer to the following document:

👉 Final Report

✍️ Reflections

“This was my very first experience using ROS and Linux, and I struggled a lot at first.
I had no one to ask, and no one on the team was familiar with ROS. So I had to learn and build almost everything from scratch.
During the final month, I worked almost every night—sleeping very little—but I poured my heart into this project.
Looking back, this is the project that made me choose robotics as my career path.”