Project AURA - Automated Urban Restaurant Assistant

Team Members

E/21/024, AMARANGA S.G.I. - Email
E/21/113, DISSANAYAKE H.G.K.V.D.C. - Email
E/21/245, MADHUSHAN S.K.A.K. - Email
E/21/407, THENNAKOON T.M.I.I.C. - Email

Introduction

In the modern hospitality industry, customers often face delays in ordering, difficulties in communicating with staff due to language barriers, and a lack of engaging entertainment while waiting. AURA (Automated Urban Restaurant Assistant) addresses these issues by introducing a smart, interactive table-top robot companion.

Unlike standard digital kiosks, AURA utilizes Social Robotics principles—employing active face tracking, voice interaction, and ambient lighting control to create a “living” digital concierge. It streamlines the ordering process, entertains customers, and assists staff by automating repetitive tasks.

Key Features

Interactive Robotics: Pan & Tilt head movement with active face tracking and voice interaction.
Autonomous Ordering: Intuitive touch-based ordering system with multi-language support.
Smart Environment: User-controlled ambient RGB lighting (Moods: Dining, Reading, Party).
Wireless & Portable: Battery-powered design requiring no table wiring (Zero Infrastructure Cost).

Solution Architecture

High-Level Overview

The system consists of three main components:

The AURA Robot Nodes: Placed at each table, handling user interaction and sensing.
The MQTT Broker: Managing real-time communication between robots and the server.
The Central Server: Handling order processing, kitchen display updates, and database management.

System Architecture

Hardware & Software Designs

Hardware Specifications

The robot is built around high-performance embedded computing to support AI tasks.

Component	Specification	Purpose
Main Controller	Raspberry Pi 4 Model B (2GB)	Core processing, AI, & UI rendering
Display	7-inch Capacitive Touch Screen	User Interface for ordering & games
Camera	Raspberry Pi Camera Module V2	Face tracking & QR scanning
Actuators	2x MG996R Servo Motors	Pan & Tilt mechanism for head movement
Sensors	PIR Motion Sensor	Presence detection & auto-wake
Power	Li-ion Battery Pack (10,000mAh)	Portable power source
Audio	USB Mic & Speaker	Voice interaction & alerts

Software Stack

Robot Frontend: Python (PyQt/Kivy) for the touch interface.
Robot Logic: OpenCV for face tracking, GPIO Zero for servo control.
Communication: MQTT Protocol (Paho-MQTT) for lightweight messaging.
Central Server: Flask (Python) web server.
Database: SQLite for managing menu items and transaction history.

Data Flow

Input: User interacts via touch or voice. Camera tracks user face.
Processing: Raspberry Pi processes inputs and generates MQTT payloads.
Transmission: Data sent over Wi-Fi to the Central Server.
Action: Kitchen Display shows the order; Robot updates UI/Lighting.

Testing

To ensure reliability in a chaotic restaurant environment, we employ a three-tier testing strategy:

Unit Testing: Individual testing of Servo mechanisms, Camera feed, and UI components.
Integration Testing: Verifying MQTT message delivery latency and reliability between Robot and Server.
User Acceptance Testing (UAT): Real-world trials in a café environment to assess battery life and user interaction flow.

Detailed Budget

Below is the estimated budget for a single AURA unit based on current component prices.

Item	Quantity	Unit Cost (LKR)	Total (LKR)
Raspberry Pi 4 Model B (2GB)	1	18,500	18,500
7-inch Capacitive Touch Screen	1	16,000	16,000
Raspberry Pi Camera Module V2	1	4,500	4,500
MG996R Servo Motors	2	1,200	2,400
10,000mAh Li-ion Battery Pack	1	6,500	6,500
USB Microphone & Speaker	1	2,500	2,500
PIR Motion Sensor	1	450	450
3D Printing Filament (PLA)	1	3,000	3,000
Miscellaneous (Wires, Screws)	-	1,500	1,500
Grand Total			55,350 LKR