Eight years ago we have designed and implemented an initial prototype of an agent-based glider pilot support system (the GliderAgent). Our aim was to validate correctness of the initial assumption that an agent-based system, combined with sensor data, can help pilots in various situations that occur during a flight. For instance, the GliderAgent was capable of detecting certain dangers, warn the pilot and, autonomously, send notification(s) to the ground station. Due to the continuous and rapid development of mobile technologies and sensors, our initial prototype has evolved. First, we moved the system, from an emulated environment, to real devices. Second, a semantic-rule-based decision making system was integrated with the GliderAgent, to analyze feeds from sensors (altitude, temperature, blood pressure, etc.) and, based on received information, to trigger appropriate behaviors. The result of our work provided a foundations for development of a general-purpose, cyber-physical system framework, which will be described in this paper. Furthermore, the developed system illustrates an interesting approach to integration of heterogeneous IoT devices (potentially also IoT platforms).

中文翻译：

---

网络物理系统建模——GliderAgent 3.0 视角

八年前，我们设计并实现了基于代理的滑翔机飞行员支持系统（GliderAgent）的初始原型。我们的目标是验证最初假设的正确性，即基于代理的系统与传感器数据相结合，可以在飞行过程中发生的各种情况下帮助飞行员。例如，GliderAgent 能够检测某些危险、警告飞行员并自动向地面站发送通知。由于移动技术和传感器的持续快速发展，我们的初始原型已经发展。首先，我们将系统从模拟环境转移到真实设备上。其次，基于语义规则的决策制定系统与 GliderAgent 集成，以分析来自传感器（高度、温度、血压等）的信息，并根据收到的信息，触发适当的行为。我们的工作结果为开发通用的网络物理系统框架奠定了基础，本文将对此进行描述。此外，开发的系统展示了一种有趣的集成异构物联网设备（也可能是物联网平台）的方法。