XBOT
Experimental AI-Powered Autonomous Robotics Platform
"Exploring the intersection of artificial intelligence, robotics, perception and natural-language interaction to create a more intelligent physical world."
Where AI Meets the Physical World
XBOT provides an open physical research framework exploring how computational intelligence interacts with physical environments through continuous feedback loops.
Sensors
Captures physical environmental inputs via camera, ultrasonic, PIR, IMU & mic.
Perception
Converts raw signals into visual features, spatial distances & audio streams.
AI / LLM
Processes contextual information using edge rules and neural language models.
Decision
Determines optimal locomotion paths, obstacle maneuvers, or spoken responses.
Action
Executes physical motor signals, voice synthesizer output, or status displays.
Environment
Physical world state changes as the robot interacts and moves within space.
A Smarter Robot for a Smarter Future
Modular physical architecture designed for experimental robotics, edge computing, and embodied artificial intelligence.
AI & Natural Language
Interfaces with local edge models and cloud language APIs for natural conversational interaction and task comprehension.
Computer Vision
Onboard optical sensor stream for visual object recognition, spatial tracking, and visual-spatial scene parsing.
Autonomous Navigation
Differential drive kinematics paired with ultrasonic obstacle avoidance for autonomous spatial exploration.
Environmental Sensing
Multi-sensor array integrating ultrasonic distance, PIR motion detection, and 6-axis IMU orientation sensing.
Motor Control
Precision PWM motor driver configuration enabling exact velocity, steering, and trajectory execution.
Voice Interaction
Microphone capture and audio speaker module enabling real-time spoken natural-language dialogue.
Modular Architecture
Open physical chassis engineered for rapid hardware iteration, sensor expansion, and custom sensor additions.
Research Platform
A proven hardware foundation for investigating multimodal AI, Vision-Language Models, and STEM educational robotics.
System Architecture
A 5-layer physical-digital control stack bridging low-level sensor feedback with high-level cognitive reasoning.
Technical Specifications
Verified physical component configuration for the current XBOT research prototype.
| Subsystem | Hardware / Software Specification |
|---|---|
| Microcontroller Subsystem | ESP32 / Arduino-class dual-core RISC microcontroller Current Prototype |
| Optical Vision Module | Onboard CMOS camera sensor for visual capture and scene streaming |
| Spatial & Motion Sensing | HC-SR04 ultrasonic distance transceiver + PIR passive infrared motion detector |
| Kinematic Guidance (IMU) | 6-axis Inertial Measurement Unit (Accelerometer + Gyroscope) |
| Audio & Speech Hardware | Electret/MEMS microphone array + digital audio amplifier speaker output |
| Locomotion & Actuation | Dual DC gear motors with differential drive wheel configuration |
| Power Management | High-capacity lithium / USB power distribution circuit |
| AI & Intelligence Layer | Hybrid architecture: Local edge micro-controllers + Cloud LLM / VLM API integration |
XBOT in Action
Watch the physical prototype demonstrate autonomous perception, obstacle detection, and conversational interaction.
Autonomous Navigation
Self-guided exploration of interior floor spaces using real-time sensor feedback.
Obstacle Avoidance
Instant ultrasonic proximity detection and trajectory recalculation.
Voice Interaction
Natural spoken dialogue and command parsing via audio input/output modules.
Visual Processing
Real-time optical streaming for spatial perception and visual object detection.
From XBOT to VLM-Enabled STEM Robotics
Connecting physical robotics engineering with advanced multimodal artificial intelligence in educational environments.
Bridging Embodied AI and Multimodal Educational Systems
"XBOT is an independent experimental robotics platform. The proposed doctoral research extends this technical foundation toward Vision-Language Models embedded in educational robotics and STEM learning environments."
Embedding Vision-Language Models in STEM Robotics for Active and Authentic Learning
The proposed research investigates how multimodal Vision-Language Models (VLMs) can be embedded into physical robotics kits to provide context-sensitive guidance, visual-spatial troubleshooting, and active learning support during student-led STEM projects.
Physical Learning Environments
Investigating real-time VLM interactions with physical robotics artefacts in classrooms.
Context-Sensitive Guidance
Generating pedagogical scaffolding rather than delivering direct solution steps.
Technical Benchmarking
Evaluating latency, visual accuracy, and edge-cloud inference constraints.
Authentic Evaluation
Empirical measurement of student engagement, problem-solving, and agency.
Identifying the Research Gap
Traditional STEM Robotics
Students build and troubleshoot physical robots using static manuals and trial-and-error, often encountering barriers when debugging physical wiring or mechanical misalignments.
Text-Based AI Assistants
Generative text AI can answer general programming questions but cannot directly perceive, visually inspect, or understand a student's physical robot construction.
VLM-Enabled STEM Robotics
Combines optical visual perception, spatial understanding, and conversational dialogue to provide real-time scaffolding directly tailored to the student's physical artefact.
⚠️ Current Research Limitations
- XBOT is currently an experimental engineering prototype platform.
- It has not yet been deployed as a validated educational intervention in school classrooms.
- Systematic VLM pedagogical evaluations represent proposed future doctoral work.
- Autonomous behavior is experimental and undergoing continuous refinement.
Potential Research & Application Areas
Versatile engineering foundation across academic, educational, and experimental technological fields.
STEM Education
Interactive physical robotics platforms for secondary and tertiary engineering learning.
Educational Robotics
Scaffolded learning kits combining microcontrollers, sensors, and intelligent feedback.
Embodied AI
Researching neural cognitive architectures embedded directly into physical hardware.
Computer Vision
Real-time optical object detection and spatial scene interpretation.
Autonomous Systems
Sensor fusion, trajectory planning, and obstacle avoidance algorithms.
Human-Robot Interaction
Conversational multimodal interfaces bridging human speech, vision, and robot action.
XBOT Hardware Gallery
Authentic physical photographs of the XBOT autonomous research platform prototype.
About the Researcher
Mohamed Khan Abdul Nasser
BEng (Hons) Software Engineering (UK) · International MBA
Technology lead and software engineer with over 12 years of professional experience building enterprise software platforms, cloud infrastructure, POS systems, and AI-enabled software solutions. Currently advancing research in embodied artificial intelligence and Vision-Language Models for STEM educational robotics.
CORE RESEARCH INTERESTS
Let's Build a Smarter Tomorrow
"XBOT is an ongoing experimental platform exploring embodied AI, robotics and the future of intelligent physical systems."