The VAIPOSA project aims to develop innovative methodologies for designing and implementing Positioning, Navigation, and Timing (PNT) components using AI and machine learning (ML), while ensuring safety, robustness, and resilience. It employs a realistic road environment simulator where an autonomous agent with an AI-based PNT system must adapt to changing conditions. The project enhances traditional PNT algorithms with AI to improve performance, particularly in degraded conditions. A key aspect is the integration of a "safety cage" architecture to monitor AI outputs. The project utilizes the VIVAS methodology for system-level verification and validation (V&V), focusing on automatic test-case generation with formal methods to ensure comprehensive coverage and safe, reliable autonomous navigation in diverse scenarios.