GUARA Platform — Modular Edge AI and Human-State Intelligence for Safety-Critical Operations

GUARA Platform is developing a modular, software-first Human-State Intelligence architecture for safety-critical operations.

Rather than functioning solely as a conventional Driver Monitoring System (DMS), GUARA combines computer vision, ocular and behavioral indicators, Edge AI and temporal signal analysis to interpret how human-state indicators evolve over time.

The platform is designed around local processing, reducing dependence on continuous cloud connectivity and supporting potential integration into different hardware platforms and operational systems.

Core technical capabilities under development include:

• Computer vision and facial/ocular signal analysis.

• Temporal interpretation of fatigue, drowsiness, distraction and attention-related indicators.

• Edge AI processing on embedded and mobile devices.

• Modular software architecture designed for reuse across operational environments.

• Local operation with reduced dependence on continuous cloud connectivity.

• Laboratory testing, technical validation and software integration.

Development and validation status

GUARA has functional prototypes developed and tested in laboratory environments, including PC-based, Raspberry Pi-based and Android implementations.

These prototypes provide an initial technical foundation. Industrial field validation, broader hardware interoperability and production-level deployment remain future development objectives.

Potential application areas

Professional Mobility is GUARA's first reference vertical, targeting professional transportation, logistics and fleet operations.

The broader platform architecture is being developed for potential applications in mining, industrial machinery, manufacturing, remote operations and human supervision of safety-critical processes.

Technology integration and collaboration

GUARA's software-first approach is intended to support several future collaboration models, including modular software engine licensing, SDK/API integration into third-party products, and complete software solutions where an integrated platform is required.

We are interested in technical exchanges, independent validation, integration studies and collaborative R&D with industrial companies, technology integrators and research institutions.

Our goal is to develop a reusable Human-State Intelligence software layer that can complement existing industrial and operational systems while supporting privacy-conscious, locally processed human-state analysis.