ROS2-Based Low-Cost Mobile Robot for Educational Assistance with Reactive Navigation and Semantic-Cached Language Processing

ROS2-based mobile assistant robot for educational environments, integrating offline NLP in Spanish, YOLOv8n object detection, and reactive navigation on Raspberry Pi 4B.

📄 Paper: ROS2-Based Low-Cost Mobile Robot for Educational Assistance with Reactive Navigation and Semantic-Cached Language Processing — Submitted to Robotics (MDPI), 2026.
👥 Authors: Sebastián Alexis Aucapiña, Nataly Cecilia Benalcázar, José Varela-Aldás and Ramiro Isa-Jara
🏛️ Institution: Universidad Tecnológica Indoamérica / ESPOCH — Ecuador

---

Overview

This project presents a low-cost educational mobile robot that combines autonomous reactive navigation with offline natural language processing in Spanish. The system runs entirely on a Raspberry Pi 4B using ROS2 Humble as middleware, making it accessible for resource-constrained educational environments such as public schools or universities in Latin America.

The robot operates in two modes:

• Education Mode — answers academic questions via voice interaction using a hybrid NLP architecture
• Navigation Mode — navigates autonomously using visual and ultrasonic perception

Total hardware cost: under $250 USD.

---

System Architecture

The system is organized into four functional layers:

┌─────────────────────────────────────────┐
│         Application Layer               │
│     (Mode Orchestrator - ROS2 Node)     │
├─────────────────────────────────────────┤
│         Processing Layer                │
│  Vision | Speech/NLP | Navigation       │
├─────────────────────────────────────────┤
│         Abstraction Layer               │
│     (ROS2 Interface Nodes)              │
├─────────────────────────────────────────┤
│         Physical Layer                  │
│  Raspberry Pi 4B | Sensors | Actuators  │
└─────────────────────────────────────────┘

ROS2 Nodes

Node	Package	Function
usb_camera_node	robot_vision	Real-time image acquisition
object_detector_node	robot_vision	YOLOv8n-based object detection
object_follower_node	robot_vision	Motion command generation
ultrasonic_node	robot_navigation	Obstacle distance measurement
imu_odometry_node	robot_navigation	Orientation estimation (MPU6050)
motor_controller_node	robot_navigation	PWM motor control (TB6612FNG)
voice_node	robot_voice	Speech recognition + TTS + NLP orchestration

---

Hardware

Component	Specification
Computing unit	Raspberry Pi 4B — 4GB RAM
OS	Ubuntu 22.04 Server (ARM64)
Motors	N20 DC motors — differential drive
Motor driver	TB6612FNG
IMU	MPU6050
Camera	USB Camera (Logitech C270)
Ultrasonic sensor	HC-SR04
Audio amplifier	PAM8403 + 8Ω speaker
Battery	7.4V LiPo 2200mAh
Voltage regulator	LM2596 DC-DC step-down

---

Software Stack

Component	Technology
Middleware	ROS2 Humble
Object detection	YOLOv8n (Ultralytics)
Speech recognition	VOSK (offline, Latin American Spanish)
TTS Online	Edge TTS
TTS Offline	Piper TTS
Local LLM	Qwen2.5-1.5B Q4_K_M (quantized)
Cloud LLM	Trinity via OpenRouter API
Semantic cache	all-MiniLM-L6-v2 (sentence-transformers)

---

Key Results

Metric	Value
Speech recognition accuracy (quiet)	98%
Speech recognition accuracy (noisy)	98%
YOLOv8n F1-score	0.975
Door detection recall	100%
Semantic cache accuracy	100%
Semantic cache avg. latency	3.8 s
Battery life — Education mode	96 min (1.46 A)
Battery life — Navigation mode	75.6 min (1.85 A)
Total hardware cost	< $250 USD

---

NLP Architecture

The NLP module uses a three-tier hybrid architecture to balance accuracy, latency, and energy consumption:

User Query (voice)
       │
       ▼
┌─────────────────┐     HIT      ┌──────────────────┐
│  Semantic Cache │─────────────▶│  Instant Response│
│  (embeddings)   │              └──────────────────┘
└────────┬────────┘
         │ MISS
         ▼
┌─────────────────┐
│  Local Model    │  Qwen2.5-1.5B Q4_K_M (offline)
│  (Qwen)         │
└────────┬────────┘
         │ Complex query / no connectivity
         ▼
┌─────────────────┐
│  Cloud Model    │  Trinity via OpenRouter (online)
│  (Trinity)      │
└─────────────────┘

The semantic cache resolved 33.3% of queries without invoking any LLM, significantly reducing energy consumption and latency.

---

Installation

Prerequisites

• Raspberry Pi 4B with Ubuntu 22.04 Server (ARM64)
• ROS2 Humble installed
• Python 3.10+

1. Clone the repository

git clone https://github.com/SesSic/MobileRobotAssistantNLPROS2.git
cd MobileRobotAssistantNLPROS2

2. Install ROS2 dependencies

sudo apt install ros-humble-usb-cam ros-humble-cv-bridge ros-humble-sensor-msgs

3. Install Python dependencies

pip install ultralytics vosk sentence-transformers openai --break-system-packages

4. Download models (not included — large files)

Model	Destination	Source
VOSK Spanish small	model-vosk-es-small/	alphacephei.com/vosk/models — vosk-model-small-es-0.42
YOLOv8n	workspace root	auto-downloaded on first run via Ultralytics
Door detector	src/robot_vision/models/	see Dataset section below
Piper TTS	/home/<user>/piper/	github.com/rhasspy/piper

Note: After downloading Piper, update the piper_path parameter in src/robot_educativo/launch/robot_completo_launch.py to match your installation path.

5. Configure audio devices

Edit src/robot_educativo/launch/robot_completo_launch.py and set your actual ALSA device IDs:

'mic_device': 'hw:1,0',       # your microphone
'speaker_device': 'plughw:2,0' # your speaker

Run arecord -l and aplay -l on the Pi to find your device numbers.

6. Add your OpenRouter API key

Edit src/robot_voice/robot_voice/unified_processor.py and replace:

'openrouter_api_key': "YOUR_OPENROUTER_API_KEY_HERE"

7. Build and launch

colcon build
source install/setup.bash
ros2 launch robot_educativo robot_completo_launch.py

---

Repository Structure

MobileRobotAssistantNLPROS2/
├── src/
│   ├── robot_educativo/        # Main package: orchestrator + launch
│   ├── robot_navigation/       # Motor, ultrasonic, IMU nodes
│   ├── robot_vision/           # Camera, YOLO, follower nodes
│   └── robot_voice/            # Voice recognition, TTS, NLP nodes
├── scripts/                    # Utility and test scripts
├── start_robot.sh              # Convenience startup script
├── test_openrouter.py          # API connectivity test
├── test_embeddings.py          # Semantic cache test
├── test_vosk_integrado.py      # Speech recognition test
└── .gitignore

build/, install/, log/, and model files are excluded from the repository. Run colcon build to generate them locally.

---

Dataset

Door detection model trained using transfer learning on the Door Dataset from Roboflow Universe (accessed March 2026).

---

Data Availability

The experimental data supporting the results of this study are available upon reasonable request: saucapina2@indoamerica.edu.ec

---

Citation

If you use this work, please cite:

@article{aucapina2026mobile,
  title={ROS2-Based Low-Cost Mobile Robot for Educational Assistance 
         with Reactive Navigation and Semantic-Cached Language Processing},
  author={Aucapi{\~n}a, Sebasti{\'a}n Alexis and Benalc{\'a}zar, 
          Nataly Cecilia and Varela-Ald{\'a}s, Jos{\'e} and Isa-Jara, Ramiro},
  journal={Robotics},
  volume={15},
  number={7},
  pages={131},
  year={2026},
  publisher={MDPI},
  doi={10.3390/robotics15070131}
}

📄 Published paper: https://doi.org/10.3390/robotics15070131

License

This project is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license, consistent with MDPI open access policy.

---

Acknowledgments

The authors acknowledge the support of the Facultad de Ingenierías, Maestría en Robótica y Automatización Industrial, Universidad Tecnológica Indoamérica, Ambato, Ecuador.