東華大學圖書館 |

Validation Methodology for Sensor Systems on Autonomous Machines for Person Detection in Public Outdoor Areas Under Variable Environmental Conditions = = Validierungsmethodik fur Sensorsysteme an autonomen Maschinen zur Personenerkennung in Offentlichen Ausenbereichen Unter variablen Umweltbedingungen de.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Validation Methodology for Sensor Systems on Autonomous Machines for Person Detection in Public Outdoor Areas Under Variable Environmental Conditions =/
Reminder of title:	Validierungsmethodik fur Sensorsysteme an autonomen Maschinen zur Personenerkennung in Offentlichen Ausenbereichen Unter variablen Umweltbedingungen de.
Author:	Meltebrink, Christian.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2023,
Description:	107 p.
Notes:	Source: Dissertations Abstracts International, Volume: 85-11, Section: A.
Contained By:	Dissertations Abstracts International85-11A.
Subject:	Farmers. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31077785
ISBN:	9798382653884

Validation Methodology for Sensor Systems on Autonomous Machines for Person Detection in Public Outdoor Areas Under Variable Environmental Conditions = = Validierungsmethodik fur Sensorsysteme an autonomen Maschinen zur Personenerkennung in Offentlichen Ausenbereichen Unter variablen Umweltbedingungen de.
Meltebrink, Christian.

Validation Methodology for Sensor Systems on Autonomous Machines for Person Detection in Public Outdoor Areas Under Variable Environmental Conditions =Validierungsmethodik fur Sensorsysteme an autonomen Maschinen zur Personenerkennung in Offentlichen Ausenbereichen Unter variablen Umweltbedingungen de. - Ann Arbor : ProQuest Dissertations & Theses, 2023 - 107 p.

Source: Dissertations Abstracts International, Volume: 85-11, Section: A.

Thesis (Ph.D.)--Technische Universitaet Berlin (Germany), 2023.

In the development of autonomous machines, the challenge of functional safety of these machines is the capability and robustness of non-contact sensor systems for protection of persons. The detection capability of the sensor systems is strongly influenced by the existing environmental conditions in the outdoor area. For a safe operation of an autonomous machine, beyond the classical consideration of functional safety, a verification of the detection capability of the sensor system in the intended application area is the next step towards safe machines. This thesis presents a concept and the realization of a novel test method with a conceptually new sensor test stand on an agricultural farm in the public outdoor area. The focus of the developed test methods is on the creation of detection areas measured under real environmental conditions (Real Environment Detection Areas - REDAs). This creates a verification of the detection capability for each sensor system under a specific environmental condition. As an application example of an agricultural machine an autonomous feeding mixer is used in cooperation with the TUV.The sensor system is not only capable of measuring in different environmental conditions: It is furthermore capable of detecting persons in the defined area. In public areas it is not possible to exclude certain groups of people such as children. Currently, standard test targets exist either for the use in public outdoor areas or for use on autonomous systems, but not for the combination of both. For this reason criteria for such a test target are established and a test target is realized based on experience from both areas. It is the first proposal for a standardized humanoid test target for autonomous mobile machines in public outdoor environments.The newly developed REDA method identifies the strengths and weaknesses of sensor systems in an unprecedented level of detail, flexibility and variance. The comparison of the sensors' measurement results can be evaluated independently of the underlying detection principle. Furthermore different parameters of the sensor system or different evaluation algorithms such as artificial intelligence, become comparable.The newly developed test method REDA presents an important and necessary contribution for the validation of sensor systems in outdoor applications. It enables a structured and verifiable validation in complex and variable environmental conditions. This enables a safe use of non-contact sensor systems for the protection of persons on manual or autonomous machines, but also in driver assistance systems. Due to it's flexibility the presented method is useful for industry and for science. Based on this test method various further investigations of sensor systems can follow.

ISBN: 9798382653884Subjects--Topical Terms:

570534
Farmers.

Validation Methodology for Sensor Systems on Autonomous Machines for Person Detection in Public Outdoor Areas Under Variable Environmental Conditions = = Validierungsmethodik fur Sensorsysteme an autonomen Maschinen zur Personenerkennung in Offentlichen Ausenbereichen Unter variablen Umweltbedingungen de.
LDR:07131nmm a2200373 4500 001 2402961
005 20241104055829.5
006 m o d
007 cr#unu||||||||
008 251215s2023 ||||||||||||||||| ||eng d
020 $a 9798382653884
035 $a (MiAaPQ)AAI31077785
035 $a (MiAaPQ)TechBerlin_20231130318324
035 $a AAI31077785
040 $a MiAaPQ $c MiAaPQ
100 1 $a Meltebrink, Christian. $3 3773222
245 1 0 $a Validation Methodology for Sensor Systems on Autonomous Machines for Person Detection in Public Outdoor Areas Under Variable Environmental Conditions = $b Validierungsmethodik fur Sensorsysteme an autonomen Maschinen zur Personenerkennung in Offentlichen Ausenbereichen Unter variablen Umweltbedingungen de.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2023
300 $a 107 p.
500 $a Source: Dissertations Abstracts International, Volume: 85-11, Section: A.
500 $a Advisor: Weltzien, Cornelia;Ruckelshausen, Arno.
502 $a Thesis (Ph.D.)--Technische Universitaet Berlin (Germany), 2023.
520 $a In the development of autonomous machines, the challenge of functional safety of these machines is the capability and robustness of non-contact sensor systems for protection of persons. The detection capability of the sensor systems is strongly influenced by the existing environmental conditions in the outdoor area. For a safe operation of an autonomous machine, beyond the classical consideration of functional safety, a verification of the detection capability of the sensor system in the intended application area is the next step towards safe machines. This thesis presents a concept and the realization of a novel test method with a conceptually new sensor test stand on an agricultural farm in the public outdoor area. The focus of the developed test methods is on the creation of detection areas measured under real environmental conditions (Real Environment Detection Areas - REDAs). This creates a verification of the detection capability for each sensor system under a specific environmental condition. As an application example of an agricultural machine an autonomous feeding mixer is used in cooperation with the TUV.The sensor system is not only capable of measuring in different environmental conditions: It is furthermore capable of detecting persons in the defined area. In public areas it is not possible to exclude certain groups of people such as children. Currently, standard test targets exist either for the use in public outdoor areas or for use on autonomous systems, but not for the combination of both. For this reason criteria for such a test target are established and a test target is realized based on experience from both areas. It is the first proposal for a standardized humanoid test target for autonomous mobile machines in public outdoor environments.The newly developed REDA method identifies the strengths and weaknesses of sensor systems in an unprecedented level of detail, flexibility and variance. The comparison of the sensors' measurement results can be evaluated independently of the underlying detection principle. Furthermore different parameters of the sensor system or different evaluation algorithms such as artificial intelligence, become comparable.The newly developed test method REDA presents an important and necessary contribution for the validation of sensor systems in outdoor applications. It enables a structured and verifiable validation in complex and variable environmental conditions. This enables a safe use of non-contact sensor systems for the protection of persons on manual or autonomous machines, but also in driver assistance systems. Due to it's flexibility the presented method is useful for industry and for science. Based on this test method various further investigations of sensor systems can follow.
520 $a Bei der Entwicklung autonomer Maschinen ist die Herausforderung der funktionalen Sicherheit dieser Maschinen die Funktionalitat und Robustheit von beruhrungslosen Sensorsystemen zum Schutz des Menschen. Die Detektionsfahigkeit der Sensorsysteme wird unter den vorhandenen Umgebungsbedingungen im Ausenbereich aufgrund der variablen Storeinflusse stark beeinflusst. Fur einen sicheren Betrieb einer autonomer Maschine ist uber der klassischen Betrachtung der funktionalen Sicherheit hinaus ein Nachweise der Detektionsfahigkeit des Sensorsystems in den geplanten Einsatzgebiet der nachste Schritt zu sicheren Maschinen. Diese Thesis prasentiert ein Konzept und die Realisierung eines neuartigen Testverfahrens mit einem konzeptionell neuen Sensorprufstandes auf einem landwirtschaftlichen Betrieb im offentlichen Ausenbereich. Der Fokus der entwickelten Testmethoden liegt auf der Erstellung von Detektionsflachen, die unter realen Umgebungsbedingungen gemessen werden (Real Environment Detection Areas - REDAs). Dadurch entsteht ein Nachweis der Detektionsfahigkeit fur jedes Sensorsystem unter einer bestimmten Umgebungsbedingung. Als Anwendungsbeispiel wird eine landwirtschaftliche Maschine in Form eines autonomen Futtermischwagens in Zusammenarbeit mit dem TUV verwendet.Neben dem Einsatz in den schwierigen Umgebungsbedingungen ist die Detektionsfahigkeit von Personen ein wichtiger Nachweis. Im offentlichen Bereich ist ein Ausschluss bestimmter Personengruppen, wie z.B. Kinder, nicht moglich. Aktuell existieren entweder Standard Prufkorper fur den Einsatz in offentlichen Ausenbereich oder fur den Einsatz an autonomen Systemen, aber nicht fur die Kombination von beidem. Aus diesem Grund werden Kriterien fur einen solchen Prufkorper aufgestellt und ein Prufkorper realisiert, der auf Erfahrungen aus beiden Bereichen basiert. Es ist der erste Vorschlag fur einen standardisierten humanoiden Prufkorper fur autonome mobile Maschinen im offentlichen Ausenbereich.Die neu entwickelte Methode REDA identifiziert die Starken und Schwachen von Sensorsystemen in einer bisher nicht gekannten Detailliertheit, Flexibilitat und Varianz. Der Vergleich ist unabhangig vom Messprinzip des verwendeten Sensorsystems moglich. Auserdem werden verschiedene Parameter des Sensorsystems oder verschiedene Auswertealgorithmen, z.B. kunstliche Intelligenz, vergleichbar.Die neu entwickelte Prufmethode REDA prasentiert einen wichtigen und notwendigen Beitrag fur die Validierung von Sensorsystem im Ausenbereich. Sie ermoglicht eine strukturierte und nachweisbare Validierung in komplexen und variablen Umgebungsbedingungen. Dadurch ist ein sicherer Einsatz von beruhrungslosen Sensorsystemen fur den Personenschutz an manuellen oder autonomen Maschinen, aber auch in Fahrerassistenzsystemen moglich. Aufgrund ihrer Flexibilitat ist sie fur die Industrie und fur die Wissenschaft von Nutzen. Basierend auf dieser Testmethode konnen verschiedene weitere Untersuchungen von Sensorsystemen folgen.
590 $a School code: 1119.
650 4 $a Farmers. $3 570534
650 4 $a Software. $2 gtt. $3 619355
650 4 $a Usability. $3 3681540
650 4 $a Machinery. $3 524780
650 4 $a Neural networks. $3 677449
650 4 $a Sensors. $3 3549539
650 4 $a Farms. $3 3221144
650 4 $a Automobile safety. $3 3681451
650 4 $a Controllers. $3 3559217
650 4 $a Robots. $3 529507
650 4 $a Milk production. $3 665496
650 4 $a Vehicles. $3 2145288
650 4 $a Agriculture. $3 518588
650 4 $a Automotive engineering. $3 2181195
650 4 $a Robotics. $3 519753
650 4 $a Transportation. $3 555912
690 $a 0800
690 $a 0473
690 $a 0540
690 $a 0771
690 $a 0709
710 2 $a Technische Universitaet Berlin (Germany). $3 3480527
773 0 $t Dissertations Abstracts International $g 85-11A.
790 $a 1119
791 $a Ph.D.
792 $a 2023
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31077785