Project Work Packages
The ANTIDOTE project is structured into multiple work packages (WPs), each focusing on different aspects of AI-driven cybersecurity:
WP1: Project Management
Tasks:
- Administrative and Financial Management: Oversee project progress, budget, and collaboration.
- Technical & Innovation Management: Ensure technical excellence and innovation.
- Data Management and Legal Compliance: Develop a data management plan and ensure GDPR compliance.
- Diversity, Ethics, and Open Science: Ensure ethical compliance and promote diversity and open science practices.
Outcome: Quality assurance and data management plans, progress reports, mid-term and final project reports, evaluation questionnaires, and mobility declarations.
WP2: AI Attack Exploration and Threat Modelling
Tasks:
- User and Technical Requirements: Collect datasets and define user-centered requirements.
- AI Threat Modeling: Develop a taxonomy for adversarial AI attacks in smart healthcare.
- ANTIDOTE Framework Design: Define the high-level architecture and requirements.
Outcome: Adversarial AI threat model and ANTIDOTE framework architecture.
WP3: Defensive AI for Secure Smart Healthcare
Tasks:
- Federated Learning for Medical Data: Optimize privacy and security in medical data analysis.
- AI-Powered Malware Detection: Detect malware in medical IoT devices using deep learning.
- AI-Powered Intrusion Detection: Develop an IDS for medical network security.
Outcome: Reports on AI solutions for medical data analysis and healthcare security.
WP4: Adversarial AI for Attacking Smart Healthcare
Tasks:
- Poisoning Medical Data: Generate adversarial examples for medical data analysis.
- Bypassing Malware Detection: Develop adversarial techniques to evade malware detection.
- Evading Intrusion Detection: Generate adversarial traffic to bypass IDS.
Outcome: Reports on adversarial AI methods for attacking smart healthcare systems.
WP5: Robust AI for Trustworthy Smart Healthcare
Tasks:
- Robust Medical Data Analysis: Develop robust AI models for medical image analysis.
- Robust Malware Detection: Enhance malware detection robustness.
- Robust Intrusion Detection: Develop a robust IDS using explainable AI.
Outcome: Reports on robust AI methods for medical data analysis and healthcare security.
WP6: Framework Integration, Demonstration, and Evaluation
Tasks:
- Integration and Testing: Integrate components and define evaluation plans.
- Demonstrators Implementation: Implement use cases for COVID-19 diagnosis and ransomware defense.
- Results Assessment: Evaluate demonstrator results and extract lessons learned.
Outcome: Integrated ANTIDOTE framework and demonstrator implementation reports.
WP7: Dissemination, Training, and Exploitation
Tasks:
- Dissemination & Communication: Promote project results and manage branding.
- Training Activities: Develop career plans and organize training events.
- Exploitation & IPR Management: Develop an exploitation roadmap and manage IPR.