Towards 3D printed modular unmanned aerial vehicle development: The landing safety paradigm

Mitroudas, Tilemahos; Tsintotas, Konstantinos A.; Santavas, Nicholas; Psomoulis, Athanasios; Γαστεράτος, Αντώνιος

doi:10.1109/ist55454.2022.9827665

Cited by 10 publications

(3 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ease in take-off and landing that the quadrotor layout exhibits, renders it the most suitable for our application [24]. The model used in the Gazebo simulation was the one presented in [25], [26].…”

Section: A Experimental Setupmentioning

confidence: 99%

Embedded light-weight approach for safe landing in populated areas

Mitroudas¹,

Balaska²,

Psomoulis³

et al. 2023

Preprint

View full text Add to dashboard Cite

Landing safety is a challenge heavily engaging the research community recently, due to the increasing interest in applications availed by aerial vehicles. In this paper, we propose a landing safety pipeline based on state of the art object detectors and OctoMap. First, a point cloud of surface obstacles is generated, which is then inserted in an OctoMap. The unoccupied areas are identified, thus resulting to a list of safe landing points. Due to the low inference time achieved by state of the art object detectors and the efficient point cloud manipulation using OctoMap, it is feasible for our approach to deploy on low-weight embedded systems. The proposed pipeline has been evaluated in many simulation scenarios, varying in people density, number, and movement. Simulations were executed with an Nvidia Jetson Nano in the loop to confirm the pipeline's performance and robustness in a low computing power hardware. The experiments yielded promising results with a 95% success rate.

show abstract

Section: A Experimental Setupmentioning

confidence: 99%

Embedded light-weight approach for safe landing in populated areas

Mitroudas¹,

Balaska²,

Psomoulis³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…where tu edge is edge server processing time, tu fog and tu cloud represent fog and server processing time, respectively. Specifically, on the edge server side, the main task is to gather and send 3D printer images data to the fog server, which is detailed in (6):…”

Section: Hierarchical Federated Transfer Learningmentioning

confidence: 99%

“…For example, it is used to produce aerospace elements, accounting for 30% of components in aeroengines [3], [4]. Additionally, it finds applications in the manufacturing of radio-frequency components [5] and parts for unmanned aerial vehicles (UAVs) [6].…”

Section: Introductionmentioning

confidence: 99%

HFTL: Hierarchical Federated Transfer Learning for Secure and Efficient Fault Classification in Additive Manufacturing

Putra¹,

Rachmawati²,

Abisado

et al. 2023

IEEE Access

View full text Add to dashboard Cite

The technology advancement is supported by additive manufacturing industries, especially 3D printing companies, that enable fast object prototyping and development in Industry 4.0. As 3D printed products are highly adopted in various fields, the final printed product must fulfill precise requirements without any defects. Therefore, an efficient framework that simultaneously learns and detects faults during the printing process is required. Unfortunately, most state-of-the-art studies utilize a centralized approach, which is inefficient for continuous model updates. This article presents a hierarchical federated transfer learning (HFTL) framework that employs edge, fog, and cloud concepts to enable an efficient model updating mechanism. The proposed HFTL uses a well-known DL model with a new classifier to significantly reduce the distributed training process while providing high detection and classification performance. Additionally, the fog server comprehensively exploits the data collection from several edge servers and performs local training. The extensive simulation results indicate that the performance of the proposed HFTL is more efficient, with 24% faster training time, effectively detects flaws in 3D printing products with 45% accuracy, and has a 59% F1-score improvement in non-IID data distribution compared to traditional FL architecture.

show abstract