info@wasabiproject.eu

CognitEye - Company and Team

Is a deep-tech company with expertise in AI and Assistive Technologies.

⇒  We are a multidisciplinary team of dedicated engineers and scientists.

Dr. Ehsan D. Farahani

Dr. Ehsan D. Farahani

  • Team manager, system design engineer
  • Data Scientist
    Amirhosein Baratai

    Amirhosein Baratai

    • Software engineer
    • Full Stack and Android developer
    Hosna Rohani

    Hosna Rohani

    • AI team leader, AI developer
    • Computer vision specialist

    EXPERIMENT OVERVIEW

    OVERVIEW

    Developing an AI-based voice assistant system for electronic assembly and installation tasks, specifically targeting Social Workplaces.

    1. Conversational AI for Task Execution
    2. step-by-step guidance

    CHALLENGES

    The challenges in social workplaces for assembly process

    • Cognitive and language barriers in assembly processes
    • High error rates and inefficiencies
    • Onboarding of new employees

    OBJECTIVES

    Objective 1: Implement conversational guidance using NLP
    Objective 2: Provide Step-by-Step Task Guidance
    Objective 3: Develop an Open-Source Solution Using OVOS

    EXPERIMENT IMPACT

    EXPECTED RESULTS (KPIs)

    • Ensure 95% accuracy in voice command recognition and response during testing.
    • Task completion time reduction of new employees – target: 20% faster assembly times compared to traditional method
    • Percentage of worker satisfaction and inclusion – target: 20% improvement in satisfaction of workers.

    AS-IS SITUATION

    • Read assembly instructions for each step of assembly task
    • Task completion is time consuming
    • Training new employees is difficult

    TO-BE SITUATION

    • Reduced cognitive load
    • Minimized errors and increased efficiency
    • Agile onboarding
    • Improved worker satisfaction and inclusion

    EXPERIMENT WORK PLAN

    WP1 (M1 - M3): Project Setup, and System Design

    Experiment work plan
    Task(s) description:
    • Taks 1: Define detailed system requirements with Bewel, and getting assembly process data from them.
    • Taks 2: Design the system architecture, including conversational AI, NLP, Docker orchestration, and COALA app integration.
    • Taks 3: Create an Intellectual Property Rights (IPR) plan (submitted in Month 2).
    Main results:
    1. Clear project requirements, objectives, and system design.
    2. Initial tracking and monitoring via the Experiment Handbook.
    3. Clear IPR strategy to protect the project’s innovations.
    Milestone(s): System design completed (Month 2).
    Deliverable(s):
    Experiment Handbook (V1) (Month 2).
    IPR Plan (Month 2).
    Demonstration 1 (Month 3).

    WP2 (M3 - M9): Development of the Voice Assistant and Integration

    Task(s) description:
    • Taks 1: Develop the knowledge base for task execution based on Bewel’s data.
    • Taks 2: Develop the AI-powered voice assistant using OpenVoiceOS (OVOS).
    • Taks 3: Implement NLP functionality for task guidance.
    • Taks 4: Integrate the voice assistant with the COALA Android app, using the Hivemind2 service for communication.
    Main results:
    1. Fully functional conversational AI for step-by-step task guidance.
    2. Integrated solution with a user-friendly interface (COALA app).
    Milestone(s):
    AI assistant prototype completed (Month 6). COALA app integration completed (Month 7).
    Deliverable(s):
    Experiment Handbook (Version 2) (Month 6).
    Demonstration 2 (Month 5).
    Demonstration 3 (Month 7).

      WP3 (M10 - M11): Testing in Manufacturing Environment

      Task(s) description: (will be done by major contribution of Digitalis)
      • Taks 1: Deploy the voice assistant solution in Bewel’s assembly lines for real-world testing.
      • Taks 2: Collect and analyze feedback from workers in assembly line.
      • Taks 3: Conduct validation to assess system performance and usability.
      Main results:
      1. Validated AI voice assistant that meets the needs of social workplace employees.
      2. Improved system performance and usability based on real-world testing.
      Milestone(s):
      Field testing completed (Month 10).
      Deliverable(s):
      Demonstration 4 (Month 9).
      Feedback analysis and validation report (Month 10).

      WP4 (M11 - M12): Final Optimization, Reporting, and Project Closure

      Task(s) description:

      • Taks 1: Final optimization of the AI system based on feedback and validation results.
      • Taks 2: Complete the final Experiment Handbook (Version 3), project report and cost statement, and final system demonstration (Month 12).

      Main results:

      1. Fully optimized, ready-to-deploy AI voice assistant for electronic assembly tasks.
      2. Final documentation capturing all aspects of the project, from development to deployment.

      Milestone(s): Project completion and final demonstration (Month 12).

      Deliverable(s):* Final Experiment Handbook (Version 3) (Month 12).* Demonstration 5 (Month 12).* Final report and cost statement (Month 12).

      BACK TO EXPERIMENTS