LuxAI QTrobot Robots – Intelligent Support for Education, Therapy, and Artificial Intelligence Research
Social robots are increasingly finding their way into schools, child development centers, research laboratories, and technical universities. Their role is not to perform physical labor, but to interact with humans: speaking, listening, recognizing gestures, expressing emotions, and implementing pre-programmed educational scenarios.
One manufacturer specializing in this field is the Luxembourg-based company LuxAI, creator of the QTrobot family of humanoid robots.
LuxAI – Manufacturer of Social Robots
LuxAI is a technology company originating from the University of Luxembourg. The company was spun out of the university in April 2016 by an interdisciplinary team including researchers, artificial intelligence and robotics specialists, healthcare professionals, and child development experts. The company develops systems using artificial intelligence and social robotics to support education, research on human-robot interaction, and working with individuals with special educational needs.
LuxAI's main product is QTrobot – a humanoid social robot designed to communicate with humans using:
speech,
on-screen facial expressions,
head and arm movements,
sound,
camera images,
voice recognition, facial recognition, gesture recognition, and body position recognition.
The manufacturer is developing QTrobot in three main directions:
as a tool for research laboratories and universities,
as an educational aid for schools and specialized centers,
as a solution supporting children's learning at home.
One QTrobot, several distinct applications
LuxAI does not offer many completely different robot designs. The core offering is a single QTrobot device family, available in several hardware versions and software packages.
The most important variants are:
QTrobot Teaching TP2 – a version intended primarily for home use;
QTrobot Teaching TP4 – a more powerful version for schools, educational centers, and therapeutic settings;
QTrobot RD-V2 i5 – a research and development platform with an Intel NUC i5 computer;
QTrobot RD-V2 i7 – a research variant with an Intel NUC i7 computer;
QTrobot RD-V2 AI@Edge – the most advanced version designed for local AI processing.
The Home, Professional, and Enterprise packages are not separate robots. They are bundles that include a specific version of QTrobot hardware, applications, tablets, educational materials, user accounts, and access to subscription services.
QTrobot Teaching TP2 – a robot for learning at home
The QTrobot Teaching TP2 is the basic educational version. It is primarily included in the QTrobot Autism Home package, designed for parents and caregivers of children requiring additional support in developing social, linguistic, cognitive, and emotional skills.
QTrobot Teaching TP2 Specifications
Quad-core ARM processor with a frequency of up to 1.2 GHz;
1 GB of RAM;
32 GB of eMMC memory;
two 10-inch Lenovo tablets;
QTrobot Parent App for parents or guardians;
QTrobot Learner App for children;
progress reporting dashboard;
the ability to create student profiles;
a two-year hardware warranty is included in the manufacturer's current package.
The package includes programs on early development, recognizing and expressing emotions, people and professions encountered in the community, and learning about clothing and independence. To use the educational portion of the system, an active software subscription is required.
The robot conducts lessons in the form of short, interactive exercises. It can issue instructions, demonstrate gestures and emotions, ask questions, and prompt the child to perform a specific activity. The curriculum can be adapted to the child's current developmental level and educational goals.
QTrobot Teaching TP4 – a solution for schools and centers
QTrobot Teaching TP4 is a more powerful educational variant included in the QTrobot Autism Professional and QTrobot Autism Enterprise packages.
The Professional version is intended for schools, clinics, special education facilities, therapists, and organizations conducting classes with larger numbers of children. The Enterprise version is designed for large organizations using at least ten robots.
QTrobot Teaching TP4 Specifications:
Quad-core ARM processor with a frequency of up to 1.8 GHz;
4 GB of RAM;
64 GB of eMMC memory;
Two 10-inch Lenovo tablets;
QTrobot Educator App;
QTrobot Learner App;
Reporting and progress monitoring dashboard;
Ability to manage accounts for teachers, therapists, and students;
Two-year hardware warranty as specified in the manufacturer's package.
The software includes programs on early development, emotions, communication, independence, and basic phonics. The Enterprise package also allows you to create your own materials in QTrobot Studio and provides additional space for content prepared by the school.
QTrobot can be used during individual sessions with a student or as an additional participant in teacher-child interactions. It doesn't replace a teacher or therapist, but performs repetitive exercises, presents instructions, and helps record the progress of the sessions. The manufacturer also provides implementation support, including device configuration, creating individual learning plans, and using progress reports.
QTrobot RD-V2 i5 – a platform for universities and laboratories
The QTrobot RD-V2 i5 is designed for research in social robotics, human-robot interaction, artificial intelligence, STEM education, and human-assisted systems.
The robot contains two collaborating computers. The first controls the device's basic components, such as the display, motors, microphones, and speakers. The second is responsible for more demanding applications, image processing, speech recognition, and running user-created programs.
QTrobot RD-V2 i5 Specifications
Main Computing Computer:
Intel NUC with a quad-core Intel Core i5 processor up to 4.4 GHz;
8 GB RAM;
128 GB SSD.
Control Computer:
Raspberry Pi 4B;
quad-core ARM processor up to 1.8 GHz;
4 GB RAM;
64 GB of eMMC memory.
Other features:
Intel RealSense D455 depth camera;
Nuitrack body and gesture tracking system;
ReSpeaker Mic Array v2.0 quad-microphone module;
Two speakers with a 2.8W stereo amplifier;
Vosk local speech recognition;
Ability to use online speech recognition services;
Two Acapela speech synthesizer voices;
ROS API [Application Programming Interface];
Python, C++, and JavaScript support;
Three-year license for the LuxAI Studio graphical environment;
USB and USB-C ports;
An adapter for connecting a monitor, Ethernet, and USB devices.
Detailed specifications for this version are published directly on the manufacturer's website.
QTrobot RD-V2 AI@Edge – Artificial Intelligence Processed Locally
The QTrobot RD-V2 AI@Edge is the most advanced research variant. Instead of an Intel NUC, it uses the NVIDIA Jetson AGX Orin platform, designed for local processing of AI models.
QTrobot AI@Edge Specifications
Main Computing Computer:
NVIDIA Jetson AGX Orin;
12-core ARM processor up to 2.2 GHz;
NVIDIA graphics card with 2048 CUDA cores and 64 Tensor cores;
Manufacturer-declared computing power of up to 275 TOPS [Tera Operations Per Second];
64 GB of LPDDR5 memory;
2 TB NVMe drive.
Control computer:
Raspberry Pi 4B;
Quad-core ARM processor up to 1.8 GHz;
4 GB RAM;
64 GB eMMC memory.
Interaction System:
Intel RealSense D455 depth camera;
NVIDIA algorithms for object and body position detection;
Nuitrack system for body and gesture tracking;
ReSpeaker four-microphone array;
local speech recognition and synthesis using NVIDIA Riva;
two speakers with a 2.8W stereo amplifier;
the ability to run local language models;
ROS API for Python, C++, and JavaScript;
DisplayPort, USB 3, and HDMI adapter.
AI@Edge's most important feature is the ability to perform all speech, image, and artificial intelligence processing directly on the robot. This can reduce the need to upload recordings or other user data to external cloud services.
QTrobot's Design and Movement Capabilities
QTrobot is approximately 63 cm tall and is designed as a robot that can be placed on a table or other stable surface. Its size allows it to interact with a seated child or adult at approximately eye level.
The robot has a metal internal structure and eight motors with metal gears. The motors control:
head rotation and tilt;
arm movements;
shoulder joints;
elbow joints.
The motors are overload-protected and transmit position, velocity, and load information. This allows you to create custom gestures, movement sequences, and more advanced scenarios using the robot's kinematics.
Screen and Emotion Display
QTrobot's head houses an 8-inch screen with a resolution of 800 x 480 pixels. The screen is non-touchscreen. It is primarily used to display the robot's face, gaze direction, eye movements, and emotions, but can also be used to display graphics and videos.
QTrobot can display over 30 facial expressions. The combination of on-screen facial expressions, head movements, arm gestures, and speech allows for more readable interactions than with a standard computer or tablet.
Camera, Microphones, and Audio
The research versions are equipped with an Intel RealSense D455 depth camera. It allows for color and spatial image capture, as well as the detection of people, body positions, gestures, and objects. The camera can be used by developers in their own applications.
The ReSpeaker Mic Array v2.0 module contains four digital microphones. It supports, among other things:
voice capture from a greater distance;
voice activity detection;
Determining the direction of sound;
Noise reduction;
Microphone beamforming;
Multi-channel audio recording.
QTrobot also has two built-in speakers with a stereo amplifier. An external microphone, speaker, or sound card can be connected to the device via USB.
Robot Software and Programming
QTrobot can be used by both programmers and those with no coding experience.
Technical users can create applications using ROS [Robot Operating System – a development environment and toolkit for robots], Python, C++, JavaScript, and available machine learning libraries. The newer QTrobot V3 documentation also describes support for Python, TypeScript/Node.js, and ROS 2.
Non-technical users can use QTrobot Studio. This graphical environment uses building blocks to create:
robot utterances;
gestures and movements;
facial expressions;
image and video presentations;
reactions to user responses;
complete interaction scenarios.
This allows programmers to collaborate with educators, psychologists, therapists, doctors, and social scientists without requiring each team member to know a programming language.
What can LuxAI robots be used for?
Educating children with special needs
QTrobot can support learning:
recognizing and naming emotions;
verbal and nonverbal communication;
vocabulary development;
understanding commands;
maintaining joint attention;
imitating movements;
solving simple problems;
independence;
behavior in social situations;
recognizing objects and activities.
The robot is primarily used with autistic children, but the manufacturer also indicates its potential use in educating students with other special educational needs.
Research
The research platform can be used in projects involving:
human-robot interaction;
social robotics;
emotion recognition;
generative artificial intelligence;
voice communication;
STEM education;
support for seniors;
healthy aging;
motor activation;
rehabilitation;
storytelling;
empathy research;
personalized robot behavior.
LuxAI presents examples of QTrobot applications in projects including education, support for seniors, physical activity, dysgraphia, teaching social-emotional skills, and building conversational robots.
Teaching Robotics and Artificial Intelligence
QTrobot can also be used as a teaching station for students of computer science, automation, robotics, psychology, pedagogy, and artificial intelligence.
Students can learn:
robot programming;
image processing;
speech recognition;
speech synthesis;
motion control;
human-robot interaction design;
language modeling;
hardware integration with cloud services;
local AI processing.
Key Advantages of QTrobot Robots
Clear and Predictable Communication
QTrobot can always repeat instructions and gestures in a similar manner. For some students, this predictability is easier to understand than rapidly changing human interaction.
Clear Emotions and Gestures
The facial expressions displayed on the screen are simplified and deliberately clear. Combined with head and shoulder movements, they can facilitate learning to recognize emotions, gestures, and nonverbal signals.
Repeatable Exercises
The robot does not tire and can repeatedly perform the same scenario. This is beneficial for exercises requiring a high number of repetitions.
Individualized Learning
Learning programs can be tailored to the student's current skills. The reporting panel allows for recording results, tracking progress, and planning further work areas.
Combining Robot, Tablets, and Software
QTrobot is not just a physical device. In educational packages, it creates an ecosystem that includes the robot, two tablets, applications, educational content, student profiles, and reporting.
Accessible to both technical and non-technical users
Developers can use APIs, ROS, and popular programming languages, while educators and social scientists can prepare scenarios in a graphical environment.
Local Data Processing in AI@Edge
The NVIDIA Jetson AGX Orin variant allows for the execution of portions of speech recognition, image recognition, and AI models locally. This is important for projects where limiting data transfer to the cloud is a significant technical or organizational requirement.
Extensive Perception System
The combination of a depth camera, directional microphones, speech synthesis, a screen, and movable arms allows for the creation of interactions that simultaneously utilize image, sound, speech, facial expressions, and gestures.
August 6, 2026
Source: LuxAI
LuxAI, QTrobot, robot społeczny, robot humanoidalny, robot edukacyjny, robot terapeutyczny, robot dla dzieci, robot dla szkół, robot dla uczelni, robot do badań, robotyka społeczna, sztuczna inteligencja, AI, interakcja człowiek–robot, HRI, edukacja specjalna, autyzm, wsparcie terapii, rozwój kompetencji społecznych, nauka emocji, komunikacja, edukacja STEM, robot programowalny, ROS, ROS 2, Python, C++, JavaScript, QTrobot Studio, kamera głębi, Intel RealSense, rozpoznawanie mowy, synteza mowy, rozpoznawanie gestów, śledzenie sylwetki, NVIDIA Jetson, AI Edge, robot badawczy, robot dla laboratoriów, robot dla terapeutów, robot dla pedagogów, robot dla psychologów, robot interaktywny, robot do nauki programowania, Social robot, humanoid robot, educational robot, therapeutic robot, robot for children, robot for schools, robot for universities, research robot, social robotics, artificial intelligence, AI, human-robot interaction, HRI, special education, autism, therapy support, development of social competences, learning emotions, communication, STEM education, programmable robot, depth camera, Intel RealSense, speech recognition, speech synthesis, gesture recognition, body tracking, NVIDIA Jetson, AI Edge, research robot, robot for laboratories, robot for therapists, robot for educators, robot for psychologists, interactive robot, robot for learning programming