東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Integration of UAV and LLM in Agricu...

Angamgari, Sudeep,

Linked to FindBook

Google Book

Amazon

博客來

Integration of UAV and LLM in Agricultural Environment /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Integration of UAV and LLM in Agricultural Environment // Sudeep Angamgari.
Author:	Angamgari, Sudeep,
Description:	1 electronic resource (52 pages)
Notes:	Source: Masters Abstracts International, Volume: 86-08.
Contained By:	Masters Abstracts International86-08.
Subject:	Large language models. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31850761
ISBN:	9798304941280

Integration of UAV and LLM in Agricultural Environment /
Angamgari, Sudeep,

Integration of UAV and LLM in Agricultural Environment /Sudeep Angamgari. - 1 electronic resource (52 pages)

Source: Masters Abstracts International, Volume: 86-08.

UAVs (Unmanned Aerial Vehicles) are autonomous or remotely controlled aircraft utilized across civilian and military sectors for tasks such as aerial photography, delivery, and surveillance. Integrating artificial intelligence has further enhanced UAV autonomy, enabling real-time data processing, adaptive decision-making, and the capability to perform complex operations without direct human intervention. Recent advances in AI and UAV hardware-such as stronger structures, improved navigation systems, extended battery life, and advanced sensors-have expanded UAV applications. In agriculture, AI-powered UAVs are used for precision farming by capturing detailed aerial imagery to monitor crop health, optimize pesticide usage, and manage irrigation systems efficiently, thus enhancing crop yields and reducing operational costs. However, existing systems often lack intuitive user interfaces and advanced features, such as natural language command processing and robust security, which are crucial for real-time operation in dynamic environments.To address these gaps, a novel UAV system was presented that leverages large language models (LLMs), specifically Generative AI, integrated with a secure voice-command interface to enhance responsiveness and operational efficiency. Unlike previous approaches that rely on traditional AI modules or basic LLM applications, this system emphasizes secure, optimized UAV operation. The design consists of a four-stage process: two-factor user authentication, streamlined command processing, dynamic threshold-based operational time optimization, and enhanced scene understanding with YOLOv8.The system was evaluated in a simulated agricultural environment created in Unreal Engine and tested using AirSim, allowing for robust interaction under varying environmental conditions, such as fog, wind, and fluctuating light levels. Each stage of the system was crafted to maximize task efficiency and adaptability to complex field scenarios.This evaluation showed significant improvements in command execution efficiency and scene understanding. Through a tokenization strategy, a 50.5-65.8% reduction in token usage was observed, effectively reducing task completion time, lowering API usage costs, and minimizing user intervention. Google Speech-to-Text provided high-accuracy voice recognition across diverse accents and noisy backgrounds, enhancing command interpretation. YOLOv8 integration enabled precise object detection for crop inspection, maintaining high accuracy under clear skies and adapting effectively to fog and crosswinds. Wind tests further indicated that high headwinds could extend task completion by up to 9 seconds, demonstrating the influence of environmental factors on UAV performance. These results validate this system's adaptability, efficiency, and robustness, highlighting its suitability for agricultural applications.

English

ISBN: 9798304941280Subjects--Topical Terms:

3775019
Large language models.
Subjects--Index Terms:

Unmanned Aerial Vehicles

Integration of UAV and LLM in Agricultural Environment /
LDR:04240nmm a22004213i 4500 001 2436229
005 20251021112918.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798304941280
035 $a (MiAaPQD)AAI31850761
035 $a (MiAaPQD)Purdue28030679
035 $a AAI31850761
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Angamgari, Sudeep, $e author. $3 3826821
245 1 0 $a Integration of UAV and LLM in Agricultural Environment / $c Sudeep Angamgari.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (52 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 86-08.
500 $a Advisors: Raja, Ashok Vardhan Committee members: Tu, Manghui Michael; Asrari, Arash.
502 $b M.S. $c Purdue University $d 2024.
520 $a UAVs (Unmanned Aerial Vehicles) are autonomous or remotely controlled aircraft utilized across civilian and military sectors for tasks such as aerial photography, delivery, and surveillance. Integrating artificial intelligence has further enhanced UAV autonomy, enabling real-time data processing, adaptive decision-making, and the capability to perform complex operations without direct human intervention. Recent advances in AI and UAV hardware-such as stronger structures, improved navigation systems, extended battery life, and advanced sensors-have expanded UAV applications. In agriculture, AI-powered UAVs are used for precision farming by capturing detailed aerial imagery to monitor crop health, optimize pesticide usage, and manage irrigation systems efficiently, thus enhancing crop yields and reducing operational costs. However, existing systems often lack intuitive user interfaces and advanced features, such as natural language command processing and robust security, which are crucial for real-time operation in dynamic environments.To address these gaps, a novel UAV system was presented that leverages large language models (LLMs), specifically Generative AI, integrated with a secure voice-command interface to enhance responsiveness and operational efficiency. Unlike previous approaches that rely on traditional AI modules or basic LLM applications, this system emphasizes secure, optimized UAV operation. The design consists of a four-stage process: two-factor user authentication, streamlined command processing, dynamic threshold-based operational time optimization, and enhanced scene understanding with YOLOv8.The system was evaluated in a simulated agricultural environment created in Unreal Engine and tested using AirSim, allowing for robust interaction under varying environmental conditions, such as fog, wind, and fluctuating light levels. Each stage of the system was crafted to maximize task efficiency and adaptability to complex field scenarios.This evaluation showed significant improvements in command execution efficiency and scene understanding. Through a tokenization strategy, a 50.5-65.8% reduction in token usage was observed, effectively reducing task completion time, lowering API usage costs, and minimizing user intervention. Google Speech-to-Text provided high-accuracy voice recognition across diverse accents and noisy backgrounds, enhancing command interpretation. YOLOv8 integration enabled precise object detection for crop inspection, maintaining high accuracy under clear skies and adapting effectively to fog and crosswinds. Wind tests further indicated that high headwinds could extend task completion by up to 9 seconds, demonstrating the influence of environmental factors on UAV performance. These results validate this system's adaptability, efficiency, and robustness, highlighting its suitability for agricultural applications.
546 $a English
590 $a School code: 0183
650 4 $a Large language models. $3 3775019
650 4 $a Decision making. $3 517204
650 4 $a Neural networks. $3 677449
650 4 $a Computer engineering. $3 621879
650 4 $a Aeronomy. $3 2102064
650 4 $a Information technology. $3 532993
653 $a Unmanned Aerial Vehicles
653 $a Aircraft
690 $a 0800
690 $a 0367
690 $a 0489
690 $a 0464
710 2 $a Purdue University. $e degree granting institution. $3 3759105
720 1 $a Raja, Ashok Vardhan $e degree supervisor.
773 0 $t Masters Abstracts International $g 86-08.
790 $a 0183
791 $a M.S.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31850761