Empowering CyberSecDome with Clean, Real-World IoT Data

At i46, we are thrilled to announce our participation in the CyberSecDome Open Call with our project SHIELD (Secure High-Impact Enhanced Learning Datasets). SHIELD aims to revolutionize AI-driven cybersecurity by providing clean, high-quality, and GDPR-compliant datasets derived from real-life IoT telemetry. These datasets will empower CyberSecDome’s tools for threat detection, vulnerability analysis, and proactive defense, while also supporting Open Call projects with scalable, actionable data.

Why SHIELD Matters

Cybersecurity tools are only as strong as the data they’re trained on. Synthetic or incomplete datasets often fall short in capturing the complexity of real-world threats. SHIELD bridges this gap by leveraging i46’s proprietary dataset of over 350 million records from 500+ IoT devices—including smart city infrastructure, industrial systems, and cloud servers—monitored by our compliance platform, i46.io.

Our dataset includes:

• Device telemetry (CPU, memory, temperature, uptime).

• Network metrics (open ports, firewall rules, outgoing connections).

• Security-specific attributes (CRA compliance status, kernel vulnerabilities, update statuses).

By refining this data through advanced cleaning techniques, SHIELD ensures it is optimized for AI training while retaining its richness and diversity—enabling CyberSecDome to uncover hidden vulnerabilities and improve model accuracy.

Key Innovations

Real-World Relevance
SHIELD’s datasets are sourced from active IoT deployments, offering insights into genuine device behaviors and attack surfaces. This contrasts with synthetic data, which often lacks contextual depth.

Ethical and GDPR-Compliant
The dataset is exempt from GDPR as it contains no personally identifiable information (PII). Access is restricted to authorized CyberSecDome members, with strict protocols to prevent misuse.

Scalable Impact
The project delivers 260+ million cleaned records in 5 weeks, support for 3+ AI models (pen-testing, risk assessment, threat detection), and seamless integration with CyberSecDome’s Central Threat Detection Engine (TDE) and Dynamic Risk Assessment (DRA) tools.

Collaboration for Stronger Cybersecurity

SHIELD is a collaborative effort between i46 and the CyberSecDome consortium. Through iterative feedback and alignment with CyberSecDome’s AI frameworks, we will ensure the datasets meet exacting standards for usability, flexibility, and transparency. The datasets are cleaned, labeled, and validated for immediate integration while retaining “less obvious” parameters to enable future AI innovation. Every dataset is fully documented with metadata and usage guidelines to ensure transparency.