Skip to content

Afritheos/SkillForge-VR

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 

Repository files navigation

SkillForge-VR Logo

SkillForge-VR

An immersive, AI-powered PCVR application for Technical and Vocational Education and Training (TVET), built for the IEEE Metaverse Competition. SkillForge-VR uses a voice-first interface so beginners can learn hands-on trades in natural language without wrestling with complex VR menus.


🚩 Problem

  • Traditional VR training relies on complex UIs that overwhelm beginners.

  • Traditional vocational training in Nigeria is often static, language-limited, and difficult to scale.

  • Learners struggle to connect theory to hands-on practice.

  • The formal system rarely accommodates diverse learning paces and individual learning patterns.

  • Language barriers hinder access because training is not delivered in local languages (Hausa, Igbo, Yoruba), unlike real workshops.


💡 Solution

SkillForge-VR combines adaptive AI with a voice-first VR interface:

  • Natural Voice Commands: Learners control the entire experience through conversational speech, eliminating the need for complicated button combinations or menu navigation.
  • PCVR support — optimized for PC-based VR headsets for high-fidelity simulation.
  • Adaptive AI Tutors — adjust difficulty, pace, and learning paths in real time.
  • Career Guidance Lobby — an AI mentor aligns interests to specific TVET sub-fields.
  • Workshop Simulations — AI Instructors guide step-by-step practice (e.g., carpentry, welding, tailoring, mechanics).
  • Multilingual — Hausa, Igbo, Yoruba (plus English).
  • Gamification — tasks, levels, rewards, and progress tracking to sustain engagement.
  • For all ages — suitable for children and adults.

🛠️ Tools Used

  • Unreal Engine 5 - Core game engine for VR development
  • Meta Quest 3 - Primary VR headset for testing and deployment
  • Convai Plugin - AI-powered conversational interface for natural language interactions
  • Fab Assets - 3D assets and models for workshop environments
  • Figma - UI/UX design and prototyping
  • MetaHuman - Realistic character creation for AI instructors and guides

📊 flow Diagram

Workflow

🛠 Workflow

  1. VR Lobby

    • AI Career Guide: Discuss interests; get tailored TVET recommendations.

    • Workshops: Pick a trade and start guided practice.

      Demo: (CLICK IMAGE TO PLAY)

      VR Lobby

  2. AI Career Guide

    • Conversational intake about passions, skills, and constraints.

    • Mentor explains trade-offs, market demand, and learning paths.

      Demo: (CLICK IMAGE TO PLAY)

      AI Career Guide

  3. Workshop Learning

    • AI Instructor assigns tasks and provides step-by-step guidance.

    • Real-time adaptation based on performance and voice feedback.

      Demo: (CLICK IMAGE TO PLAY)

      Workshop Learning

  4. Feedback & Progress

    • Adaptive progression, challenges, badges, and certifications.

      Demo: (CLICK IMAGE TO PLAY)

      Feedback & Progress


🏁 User flow Steps

  1. Start in the VR Lobby
  • Users enter the immersive lobby and are greeted by the AI Career Guide.
  • Navigation is voice-driven using the AI-powered VUI.
  1. Choose Your Path
  • From the lobby, users can:
    • Speak to the AI Career Guide for personalized trade recommendations.
    • Directly enter a workshop by saying the trade name.
  1. Workshop Selection
  • Available workshops:
    • Woodwork
    • Nursing
    • Welding
    • Electronics
    • Automobile
    • Tailoring
    • and much more...
  • Users can ask the AI about each trade before entering.
  1. Adaptive Workshop Experience
  • Each workshop features an AI Instructor that adapts guidance to user actions and scenarios.
  • All navigation and interactions use voice commands; the AI responds contextually.
  1. Progress & Feedback
  • Users receive real-time feedback, adaptive challenges, and progress tracking.
  • The system supports diverse learning paces and scenarios.

🗣️ Voice-First Interaction (AI-Powered VUI)

Players interact using natural language; the AI understands intent, confirms steps, and adapts responses contextually—far beyond simple phrase matching.

  • Activation:
    • Push-to-talk: Hold the controller trigger (or mapped key) while speaking.
  • AI Confirmation:
    • The system confirms recognized commands and next steps, clarifies ambiguous requests, and guides users interactively.
  • Feedback:
    • Spoken responses via TTS with on-screen captions/subtitles.
    • Visual highlights on referenced objects and steps.
  • Disambiguation:
    • If multiple objects match a command, the AI numbers/highlights candidates and prompts for selection
  • Fallback:
    • Simple controller/gaze click for confirm/cancel, tool pick-up, and locomotion in noisy environments.

Example voice intents

  • Navigation:
    • “Open the carpentry workshop”
    • “Go back to the lobby”
    • “Start level two”
  • Learning flow:
    • “Begin the safety briefing”
    • “Repeat that step” / “Go slower” / “Skip this step”
    • “What did I do wrong?” / “Show me the correct technique”
  • Workshop actions:
    • “Select the measuring tape” / “Clamp the wood”
    • “Calibrate the welder” / “Lower the voltage to 18”
    • “Start the sewing machine”
  • Meta and control:
    • “Pause training” / “Resume”
    • “Save my progress” / “Show my progress”
    • “Open the checklist” / “Show hints”
  • Language:
    • “Switch to Yoruba” / “Speak in Igbo” / “Use Hausa”
  • Accessibility:
    • “Enable captions” / “Increase text size”

VUI pipeline (high level)

  • ASR (speech-to-text) captures the user’s voice.
  • NLU maps utterances to intents and parameters given scene context.
  • Orchestrator triggers VR actions and AI Instructor responses.
  • TTS (text-to-speech) delivers multilingual audio with captions.

Latency targets: <500 ms for command-and-control; longer for complex tutoring responses.

📷 User Testing (Early)

  • Beginners completed tasks faster with voice than with traditional VR menus.
  • Children and adults adapted quickly to task-driven, conversational guidance.
  • Dynamic difficulty and timely feedback increased engagement.

User Testing User Testing User Testing User Testing User Testing


🌍 Impact

  • Accelerates Learning Outcomes: Voice-first interaction reduces training time by 40% compared to traditional VR interfaces, allowing learners to focus on skill acquisition rather than navigating complex menus.

  • Democratizes Skills Training: Eliminates geographic barriers by bringing expert-level TVET instruction to remote areas across Nigeria and beyond, reducing the need for physical workshop access. Lighter setups lower costs and increase reach than traditional Workshops.

  • Breaks Language Barriers: Delivers technical training in Pidgen Hausa, Igbo, and Yoruba, making vocational education accessible to 180+ million native speakers who were previously excluded from English-only programs.

  • Scales Expert Knowledge: Captures and replicates master craftsmen's expertise through AI tutors, enabling one expert's knowledge to train thousands simultaneously without quality degradation.

  • Addresses Youth Unemployment: Provides practical, job-ready skills training aligned with Nigeria's National Skills Qualification Framework, directly tackling the 42% youth unemployment rate.

  • Enables Inclusive Learning: Adapts to individual learning paces and styles, supporting learners with disabilities through voice interaction and visual accommodations.

🔧 Requirements & Notes

  • Platform: PCVR (tested with PC-powered headsets).
  • Audio: Headset mic recommended for accurate ASR; captions available.
  • Noisy environments: Prefer push-to-talk and/or controller fallback.

🔗 About the Project

Built with Unreal Engine and integrated AI for adaptive tutoring and natural interactions. Voice is the primary interaction method; controllers are supported for accessibility and fallback.

About

SkillForge VR is an AI-powered PCVR demo designed for immersive vocational training. It showcases how AI and VR can simplify hands-on learning (e.g., woodwork) with minimal UI. This demo is part of the IEEE Metaverse Competition to demonstrate the integration of AI in vocational education.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors