The increasing complexity of distributed robotics has driven the need for platforms that seamlessly integrate edge, fog, and cloud computing layers while meeting strict real-time constraints. This paper introduces BlazeAIoT, a modular multi-layer platform designed to unify distributed robotics across heterogeneous infrastructures. BlazeAIoT provides dynamic data transfer, configurable services, and integrated monitoring, while ensuring resilience, security, and programming language flexibility. The architecture leverages Kubernetes-based clusters, broker interoperability (DDS, Kafka, Redis, and ROS2), and adaptive data distribution mechanisms to optimize communication and computation across diverse environments. The proposed solution includes a multi-layer configuration service, dynamic and adaptive data bridging, and hierarchical rate limiting to handle large messages. The platform is validated through robotics scenarios involving navigation and artificial intelligence-driven large-scale message processing, demonstrating robust performance under real-time constraints. Results highlight BlazeAIoT's ability to dynamically allocate services across incomplete topologies, maintain system health, and minimize latency, making it a cost-aware, scalable solution for robotics and broader IoT applications, such as smart cities and smart factories.

翻译：分布式机器人系统的日益复杂性催生了对于能够无缝集成边缘计算、雾计算与云计算层级，同时满足严格实时性约束的平台的需求。本文提出BlazeAIoT，一种模块化多层平台，旨在统一异构基础设施上的分布式机器人系统。BlazeAIoT提供动态数据传输、可配置服务与集成化监控，同时确保系统的弹性、安全性及编程语言灵活性。该架构利用基于Kubernetes的集群、代理互操作性（DDS、Kafka、Redis与ROS2）以及自适应数据分发机制，以优化多样化环境中的通信与计算。所提出的解决方案包含多层配置服务、动态自适应数据桥接及分层速率限制机制，以处理大规模消息。平台通过涉及导航与人工智能驱动的大规模消息处理的机器人场景进行验证，展示了在实时约束下的鲁棒性能。结果突显了BlazeAIoT在不完整拓扑结构中动态分配服务、维持系统健康状态并最小化延迟的能力，使其成为适用于机器人及更广泛物联网应用（如智慧城市与智能工厂）的具备成本感知能力的可扩展解决方案。