Schlussbericht über das Forschungsprojekt KIZAM - Künstliche Intelligenz im Ziele- und Anforderungsmanagement; Teilvorhaben: Anforderungsdefinition und Anwendbarkeit aus Zulieferersicht

dc.contributor.authorSommerfeld, Tobias
dc.contributor.authorMüller, Andreas
dc.contributor.authorKostov, Nikola
dc.date.accessioned2025-08-11T11:07:23Z
dc.date.available2025-08-11T11:07:23Z
dc.date.issued2024-12-30
dc.description.abstractFahrzeugentwicklung durch den Einsatz künstlicher Intelligenz (KI) zu optimieren. Im Fokus standen drei technische Lösungsansätze: Formalisieren und Prüfen, Suchen und Finden sowie Model-Based Systems Engineering (MBSE). Im Rahmen von Formalisieren und Prüfen wurden Methoden zur strukturierten und maschinell überprüfbaren Formalisierung natürlicher Anforderungen entwickelt. Der Ansatz Suchen und Finden ermöglichte mithilfe von Knowledge Graphs eine verbesserte Wissensrepräsentation und die effiziente Identifikation relevanter Informationen. MBSE untersuchte modellbasierte Methoden, um Zusammenhänge zwischen Anforderungen und Systemstrukturen zu analysieren und Konflikte zu lösen. Die erarbeiteten Algorithmen und Softwarelösungen wurden in Demonstratoren integriert, validiert und erwiesen sich als praxistauglich für die Optimierung des Anforderungsprozesses. Trotz Herausforderungen, insbesondere bei der Systemintegration des Softwareprototypen in eine BMW Produktivumgebung und Nutzerakzeptanz, leistete das Projekt einen wichtigen Beitrag zur Effizienzsteigerung und Qualitätsverbesserung im Anforderungs-Management der Fahrzeugentwicklung.ger
dc.description.abstractThe research project KIZAM has successfully developed solutions for optimizing the Requirements Management (RM) in the vehicles development using Artificial Intelligence (AI): for formalization and verification, for searching and finding if knowledge, as also for model-Based Systems engineering (MBSE). While optimizing the formalization and verification, methods were developed for structured and automatic verification of natural requirements. The usage of Knowledge Graphs in search and find allowed an improved representation of knowledge and an efficient identification of relevant information. MBSE was used for researching of model-based methods to analyze contextual links between requirements and system structures and to solve conflicts between them. The developed algorithms and software solutions were integrated and validated in demonstrators and have been proved as practical for optimization of the requirements process. Despite challenges, especially on system integration of software prototypes into a BMW productive environment and on the user acceptance, the process achieved an important contribution on efficiency improvement and quality optimization of RM in the vehicle development.eng
dc.description.versionpublishedVersion
dc.identifier.urihttps://oa.tib.eu/renate/handle/123456789/20747
dc.identifier.urihttps://doi.org/10.34657/19764
dc.language.isoger
dc.publisherHannover : Technische Informationsbibliothek
dc.relation.affiliationSchaeffler Technologies AG & Co. KG
dc.relation.isSupplementedByhttps://kizam.de/
dc.rights.licenseCreative Commons Attribution-NonDerivs 3.0 Germany
dc.rights.urihttps://creativecommons.org/licenses/by-nd/3.0/de/
dc.subject.ddc000 | Informatik, Information und Wissen, allgemeine Werke
dc.subject.ddc600 | Technik
dc.subject.otherAnforderungsmanagementger
dc.subject.otherKünstliche Intelligenzger
dc.subject.otherFahrzeugentwicklungger
dc.subject.otherKnowledge Graphger
dc.subject.otherProduktentwicklungsprozessger
dc.subject.otherMBSEger
dc.subject.otherKIger
dc.subject.otherLLMger
dc.subject.otherAIeng
dc.subject.otherLLMeng
dc.subject.otherPEPger
dc.subject.otherartificial intelligenceeng
dc.subject.otherrequirements engineeringeng
dc.subject.otherrequirements managementeng
dc.subject.otherProduktentwicklungsprozessger
dc.subject.otherDSLger
dc.subject.otherWissensrepräsentationger
dc.subject.otherSemantische Netzwerkeger
dc.subject.otherRedundanzprüfungger
dc.subject.otherKonsistenzprüfungger
dc.subject.otherAutomotivger
dc.subject.otherdevelopment processeng
dc.subject.otherlarge language modeleng
dc.subject.sdg8
dc.titleSchlussbericht über das Forschungsprojekt KIZAM - Künstliche Intelligenz im Ziele- und Anforderungsmanagement; Teilvorhaben: Anforderungsdefinition und Anwendbarkeit aus Zulieferersichtger
dc.typeReport
dc.typeText
dcterms.event.date01.07.2021 -0.09.2024
dcterms.extent28 Seiten
dtf.funding.funderBMWE
dtf.funding.program19I21029E
dtf.funding.verbundnummer01237178
dtf.version1.0
tib.accessRightsopenAccess

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
KIZAM_Schlussbericht_öffentlich_partnerspezifisch_Schaeffler_Sommerfeld_2024.pdf
Size:
940.13 KB
Format:
Adobe Portable Document Format
Description:
The research project KIZAM has successfully developed solutions for optimizing the Requirements Management (RM) in the vehicles development using Artificial Intelligence (AI): for formalization and verification, for searching and finding if knowledge, as also for model-Based Systems engineering (MBSE). While optimizing the formalization and verification, methods were developed for structured and automatic verification of natural requirements. The usage of Knowledge Graphs in search and find allowed an improved representation of knowledge and an efficient identification of relevant information. MBSE was used for researching of model-based methods to analyze contextual links between requirements and system structures and to solve conflicts between them. The developed algorithms and software solutions were integrated and validated in demonstrators and have been proved as practical for optimization of the requirements process. Despite challenges, especially on system integration of software prototypes into a BMW productive environment and on the user acceptance, the process achieved an important contribution on efficiency improvement and quality optimization of RM in the vehicle development.