Adaptive Event Driven Framework for Real Time Multi-Agent Missions

Abstract

A Ground Control Station (GCS) is an essential element to supervise and control autonomous vehicles performing complex missions in real time. In the new era of Internet of Things, where systems are highly connected, these missions demand enormous amounts of computational power to correctly manage the coordination of all the vehicles involved. In this scope, the set of Unmanned Vehicles (UVs) included in the mission must achieve more difficult tasks everyday. As a consequence, the development of a robust, reusable and adaptable GCS framework to allow a single operator to monitor and control a team of heterogeneous agents raises a number of research and engineering challenges. In this paper we introduce an adaptive event-driven framework specially designed for GCSs involved in heterogeneous multi-agent missions that takes advantage of two features: 1) it allows the GCS to add or remove both actual or simulated agents in real time, changing the number or types of monitored agents, and 2) from a software design perspective, the graphical user interface dynamically changes its view in order to minimize operators fatigue and mental workload, facilitating the success of the mission in such complex environments. We also show one of the tests performed with the adaptive framework, where after observing how a real UV deployed in a water surface performs successfully a set of previously planned trajectories, we will see how a simulated UV joins the mission in order to fulfill a leader-follower maneuver.