EROS

EROS addresses a key challenge for future Smart Cities: deploying robotic assistants capable of delivering secure, reliable, and human-centric services in environments such as municipal buildings, hospitals, transport hubs, and hotels.

Current robotic solutions often rely heavily on cloud-based AI, leading to latency issues, reduced interaction quality, and increased privacy risks—especially when handling sensitive healthcare or hospitality data. Moreover, many service robots lack domain-specific datasets required for natural and adaptive human-robot interaction.

EROS overcomes these limitations by integrating three core innovations. First, it leverages EdgeMind technology to deploy lightweight Large Language Models (LLMs) directly on robots and edge devices, enabling real-time multilingual speech and text processing without dependency on distant cloud infrastructure. Second, it incorporates HOSPIT-LLM, a multimodal dataset combining speech, gestures, and video captured in hospitality and healthcare contexts, allowing robots to perform receptionist-style services such as navigation assistance, information provision, and check-in support. Third, EROS embeds Post-Quantum Cryptography (PQC) and AI-based Key Management Systems to ensure quantum-safe, trusted communication between robots, edge platforms, and Smart City infrastructures.

The solution will be demonstrated on the OTE 5G Standalone testbed in Athens, exploiting edge cloud, MEC, and network slicing capabilities to validate ultra-low latency, high reliability, and secure communications.

By combining robotics, edge AI, and quantum-safe technologies, EROS strengthens Europe’s leadership in secure, human-centric AI and supports the transition toward Industry 5.0 Smart Cities.