INTEGRATING ARDUPILOT SITL FOR DRONE BEHAVIOUR SIMULATION IN GAMIFIED TRAINING

Abstract

New educational technologies, as well as the challenges of the modern world, encourage the search for ways to improve and develop training methods for specialists to operate in dangerous spaces using unmanned swarm systems. Using technologies such as Unity and ArduPilot, it is possible to create a simulated learning environment that, through gamification mechanics, provides a sense of reward and challenge while learning to interact with a swarm of drones, which requires extensive interdisciplinary expertise. The subject of the study is the interaction module of the Unity game engine and the ArduPilot SITL simulation engine. The object of the study: a software system for training specialists to operate in dangerous spaces using drone swarms. The purpose of the article is to suggest ways to integrate the Unity game engine and the ArduPilot SITL simulation engine in context of a gamification of training specialists to operate in dangerous spaces. To achieve this objective, the paper addresses the following tasks: to formulate the concept of an application that simulates missions with drone swarms in dangerous spaces and provides progression and assessment through gamification mechanics; to justify the choice of a technological stack, in particular ArduPilot, SITL, MAVLink, Unity; to describe the architecture of the Unity ↔ SITL interaction, using TCP/UDP communication, MAVLink message processing, as well as the separation of responsibilities between layers and components. As a result of the research, the architecture was substantiated, and an interaction module between Unity and ArduPilot SITL was implemented using the MAVLink protocol, providing two-way data exchange and scalable integration of multiple simulated devices within a single application. To evaluate the scalability of the system, an experiment was conducted to analyze CPU resource usage and model update time depending on the number of simulated drones. The results showed an almost linear relationship between system load and the number of drones, and no bottleneck was observed in Unity’s main thread.