Unmanned Aerial Vehicles (UAVs) and human teams are vital in Search and Rescue (SAR) missions. The inherent flexibility of UAVs increases search efficiency. The human teams guarantee the timely rescue of targets. However, research on integrating collaborative search between teams and UAVs in SAR missions is still limited. In this paper, we propose an approach to UAV-human collaboration for SAR missions to minimize the expected time of the SAR mission completion. The method mainly faces the problem of scheduling between different teams for multi-target search. Based on the objectives mentioned above, we first rasterize the post-disaster area to search the area effectively. Secondly, a Regional Collaboration Strategy (RCS) is proposed to schedule different teams rationally. Finally, we propose the Path-Greedy Multiagent Proximal Policy Optimization (GPMAPPO) algorithm which combines the Path-Greedy algorithm with the MAPPO algorithm to select the path points rationally. The experimental results demonstrate the superiority of the proposed algorithm in terms of strategy and minimizing the expected time.
With the continuous expansion of network scale, data centers are increasingly adopting Dedicated Processing Units (DPU) to offload network tasks. However, AI deep learning tasks heavily burden the host, and such computation-intensive tasks struggle on the DPU due to limited storage and constrained computing capabilities. To address this issue, Our research focuses on the NP multi-cores architecture of DPU and innovatively leverages these micro-cores, originally intended for network packet forwarding, to deploy AI neural network inference on the DPU, achieving the first full offloading of convolutional neural network inference tasks to the NP cores of the DPU. We propose an AI inference acceleration system based on soft-hard collaboration, which innovatively proposes model format self-adjustment, block sliding memory access algorithm, and multi-core parallel pipeline scheduling strategy by combining hardware characteristics. Based on these methods, we have successfully implemented a complete CNN inference process for network anomaly traffic detection and ransomware detection scenarios on Huawei DPU chips. The system can perform CNN inference on NP cores with only 2kB of running memory, effectively reducing the peak runtime memory by up to 70%, and flexibly scheduling through the deployment of integer computation paths and soft floating-point computation paths. Compared to the direct offloading method, this system achieves 17-22x acceleration on the NP cores of the DPU, even if the performance of the NP cores is only one tenth of that of the CPU in the host, it can still achieve inference latency comparable to that of the host.
We present a novel methodology of embedded systems design based on Sleptsov net (SN) formalism. We use an SN as a graphical language of concurrent programming and as a modeling language for plant specification and integrated model composition. Among the advantages, we mention the applicability of formal verification techniques for reliable embedded system design, vivid graphical language, and compatibility of the tool-chain with different classes of hardware. Composed tool-chains are supplied with our ad-hoc tools for embedded systems design on both microcontrollers and FPGAs. We develop a software SN machine for microcontrollers and a generator of Verilog programs for FPGAs. Compared to known SN machine implementations on desktop computers and GPUs, we developed an indexed sparse matrix data structure to optimize both memory usage and performance. Benchmarks on real-life Sleptsov net programs show the robustness of the approach with considerably higher performance on FPGA.
Communities have various ways of preserving culture.One of the ways to preserve Javanese traditional arts is to form the Jaran Kepang art community.This article was written based on the results of research conducted in Jetak Hamlet, Duren Village, Bandungan District from January to August 2020.The purpose of this study was to obtain an overview of the efforts of the Jetak Hamlet community to internalize cultural preservation values through traditional Javanese arts in school age children.This study uses a qualitative method.The data collection techniques used in research include: observation, in-depth interviews and documentation.In testing the validity of the data, the authors used triangulation techniques.Data analysis used interactive qualitative data analysis with the following stages: data collection, data reduction, data presentation, and conclusion or verification.The results showed that the community played a role in internalizing the values of cultural preservation through Javanese traditional arts in school age children
Popular medicine is one of the important sectors in the treatment system in a society.Before deciding to go to the available health services, the community will usually try to cure their illnesses with materials available at home.One ingredient that is widely used in popular medicine is medicinal plants.This medicinal plant is believed to have properties in curing diseases and maintaining health and fitness.This research was conducted in Bendan Duwur Urban Village in Semarang City, using qualitative methods.Observation and interview techniques were applied in this research to obtain detailed data about the knowledge of medicinal plants in popular medicine.The results of this study indicate that there are several medicinal plants that are processed into an herb commonly used in curing diseases.Sources of knowledge of medicinal plants in the community also vary, knowledge transmitted from older generation, media, and others.With the setting of a research landscape in an urban area where there are densely populated settlements, people use flower pots to grow medicinal plants and utilize the narrow and vacant land left in their environment.This research is important because although various modern health services are available, people still use their popular knowledge in maintaining their health.
