The Prometheus methodology is a set of detailed processes of specifying, designing, implementing, and testing agent-oriented software systems. An agent-oriented system consists of multiple agents, each of which has characteristics of autonomy and robustness in dynamic environments due to having a goal and plans. The typical Internet of Things (IoT) environment is dynamic, unpredictable, and unreliable, since devices are exposed to the physical world. An agent-oriented system fits into such an IoT environment because it can achieve a goal in a dynamic, unpredictable, and unreliable environment by utilizing several different combinations of multiple applicable plans and subgoals. The contribution of this paper includes introducing an agent-oriented architecture for an IoT medical application system.

This paper includes two main ideas: medical IoT applications and a multiagent-oriented software architecture. In the medical IoT application, this paper emphasizes the importance of monitoring systems for people with mental disorders. Both real-time patient monitoring with durable IoT devices and big data solutions for developing predictive capabilities are considered as essential requirements of the system. To implement the system, this paper follows the Prometheus methodology, which consists of three phases: system specification phase, architectural design phase, and detailed design phase. The paper proposes six actors (user, cloud, semantic actor, storage actor, audit actor, thing) for a medical IoT application, though these actors can be applied to different IoT applications as well.

Even though the primary intent of the scheme is to monitor mental health, it is hard to associate the system and the purpose since it is not clear which features of the system are specific to mental health monitoring. Rather, the system is applicable to general IoT applications and readers will be confused as to why the authors explain the mental health monitoring technologies in such detail. This paper reports only the overview for the architectural phase and the detailed design. Furthermore, no implementation result is presented, so it is hard to determine whether the Prometheus method fits well into the mental health monitoring system.