Trajectory planning for unmanned aerial vehicles (UAVs) in dynamic and partially observable environments becomes more complex when extended from two-dimensional to three-dimensional navigation. Although Deep Reinforcement Learning (DRL) methods have shown strong performance in 2D scenarios, their application to 3D spaces requires redesigned observation models, action representations, and safety mechanisms. This paper extends a 2D DRL-based trajectory planning framework to 3D environments using Proximal Policy Optimization (PPO), Deep Q-Network (DQN) and Deep Deterministic Policy Gradient (DDPG). UAV agents are trained to reach randomly placed 3D targets while avoiding static and dynamic obstacles using only local sensory information. The observation space combines a local 3D occupancy representation with a relative 3D goal vector, preserving partial observability and avoiding reliance on a global map. This article proposes that the simulation results demonstrate robust, collision-aware navigation and improved safety and trajectory efficiency in each one of the DRL algorithms implemented, each having positive and negative specifics.
Eduardo Machado Wullner, Arthur von Groll dos Santos, M. Vizzotto et al.· International Conference on...· 0 citations
: This paper presents the design and evaluation of a heterogeneous mobile edge node based on the STM32F103VET6 microcontroller, optimized for efficient data collection in Smart City vehicular environments. The challenge of high CPU overhead, energy consumption and latency caused by frequent write operations to mass storage via SPI is addressed through a temporary storage management architecture employing batching and double-buffering techniques with non-blocking DMA control. The system was evaluated using BLE, 802.15.4, and WiFi communication protocols under intermittent connectivity scenarios. Experimental results show that by using an optimal buffer size, the edge node reduces SPI bus usage for reading and writing data to the SD card from 38.0% to 3.1%. This significantly minimizes delays caused by frequent SD card access and also lowers the edge node’s energy consumption by 35.3%. Additionally, a 100% data recovery rate was verified during network intermittency. This research contributes to remote monitoring in smart cities using low-cost, low-power VANET networks, providing an efficient edge computing solution under strict storage and energy constraints.
L. Campoverde, M. Tropea, F. De Rango· Proceedings of the 16th Inte...· 0 citations