Publications
Researching together and creating knowledge for a growing ecosystem
Künstliche Intelligenz in der Bauwirtschaft: Grundlagen und Anwendungsfälle
Buchveröffentlichung (2024)
Dieses Fachbuch präsentiert das Potenzial Künstlicher Intelligenz (KI) im Bauwesen. Die steigende Datenmenge und die Nutzung von Hard- und Software in Bauprojekten ermöglichen projektübergreifende Analysen mithilfe von Methoden wie maschinellem Lernen und Robotik. Die Beiträge bieten einen Überblick über KI-Methoden und zeigen anhand konkreter Anwendungsfälle aus Forschung und Praxis, wie KI im Bauwesen eingesetzt wird. Dies verspricht Effizienzgewinne, Fehlerreduzierung durch Automatisierung und neue Möglichkeiten der Entscheidungsunterstützung. Das Buch richtet sich an Führungskräfte und Fachexperten der Bauwirtschaft, die mithilfe von KI nachhaltige Verbesserungen in ihren Unternehmensprozessen anstreben.
Bewertung analytischer Planungsmethoden zur Erhöhung der Prognosesicherheit von Rahmenterminplänen bei Bauprojekten
Dissertation (2023)
In Bauprojekten sind frühzeitige Prognosen der Projektdauern entscheidend, da Abweichungen erhebliche Auswirkungen auf Termine, Stakeholder-Anforderungen und Projektabläufe haben. Die Praxis zeigt jedoch Herausforderungen bei genauen Prognosen aufgrund unbekannter Informationen und systematischer Planungsfehler. Die Forschung konzentriert sich zunehmend auf analytische Planungsmethoden, um diese Fehler zu reduzieren. Die vorliegende Arbeit untersucht den Einsatz solcher Methoden in frühen Bauprojektphasen, beleuchtet Strategien zur Abweichungsreduktion und vergleicht die Prognosegenauigkeit verschiedener analytischer Ansätze. Die Ergebnisse bieten Bauherren Orientierung für die Dauerplanung und gezielte Anwendung analytischer Methoden in frühen Projektphasen.
Investigating the Impact of Lean Leadership on Construction Project Success
Part of Engineering Management Journal (2023)
Lean, crucial for optimizing construction project performance, necessitates more than tools; it demands a cultural shift. Lean Leadership emerges as pivotal, integrating Lean principles with leadership behavior for lasting change. Despite limited studies in the construction industry, this research introduces novel components of Lean Leadership. A conceptual framework explores the relationship’s impact on construction project success. An online survey of U.S. lean construction practitioners substantiates a direct, positive association between Lean Leadership and project success, emphasizing the transformative potential of effective leadership in implementing Lean principles in construction projects.
A comparative evaluation of decision trees and expert intuition to predict durations in the predesign phase
European Conference on Computing in Construction and the 40th International CIB W78 Conference (2023)
This study assesses the prediction accuracy of CatBoost and expert intuition in estimating the duration of public construction projects during the predesign phase. Using a dataset of 367 projects from New York City, the comparison involves Mean Absolute Error (MAE) and absolute preference as performance indicators. Expert intuition exhibits high outliers, whereas CatBoost provides more consistent predictions. In uncertain situations, especially, CatBoost demonstrates practical advantages over human expert estimations, highlighting its potential utility in enhancing accuracy and reliability in construction project scheduling
Combining BIM and lean methods for scheduling for verifying and communicating demolition processes in an early project stage
European Conference on Computing in Construction and the 40th International CIB W78 Conference (2023)
Demolition projects, fraught with challenges like safety regulations and diverse stakeholders, often involve complex data in various formats. This paper advocates a solution combining BIM, project data, and Lean scheduling for early project visualization and simulation, requiring minimal client effort. Illustrated through an industrial demolition project, the method enhances communication and reduces complexity. The digital model, integrating all relevant data, fosters stakeholder understanding, enabling efficient communication, quicker decision-making, and evaluation of demolition schedules. This approach proves effective when a BIM model and Lean schedule are available, streamlining processes for improved project outcomes.
Challenges in collecting and managing data for AI application in small and medium-sized construction enterprises
European Conference on Computing in Construction and the 40th International CIB W78 Conference (2023)
Demolition projects encounter challenges adopting smart tools like BIM, Big Data, and AI in construction. Despite their prevalence, integration into daily operations is limited. A case study with a German construction SME highlights data collection challenges. Though digital and data-driven, insufficient data availability and quality hinder effective AI implementation. The paper suggests implications for SMEs to harness AI benefits in the future, emphasizing the need for enhanced data practices in construction projects for successful integration of advanced technologies.
Process Analysis with an Automatic Mapping of Performance Factors Using Natural Language Processing
31st Annual Conference of the International Group for Lean Construction (2023)
In lean construction, efficient methods are crucial for utilizing data collected during process analysis. This paper aims to automate the mapping of historical performance factors to new project specifications using Natural Language Processing (NLP). The NLP model compares process descriptions, filtering the relevant performance factors and calculating durations. This supports process analysis, aiding in generating a validated construction schedule. The case study demonstrates promising results in mapping quality and prediction accuracy. The developed concept allows last planners to validate duration estimates in lean construction process planning, promoting project stability. A more detailed process description could further enhance prediction results.
Utilising Design Thinking to Refine Customer Requirements – a Case Study Using the Concrete Supply Chain as an Example
31st Annual Conference of the International Group for Lean Construction (2023)
The German concrete supply chain faces challenges due to its high participant count, resulting in complex interface management. Previous attempts at establishing a digital information chain emphasized technology over user needs, leading to limited success. This paper presents a human-centered research methodology using Design Thinking to actively involve users, aiming to reduce waste and unnecessary activities in the concrete supply chain. Focusing on the ultimate customer, the approach refines customer requirements and offers a model for Design Thinking application in addressing industry challenges.
International assessment of artificial intelligence applications in the AEC sector
Part of ECPPM 2022 - eWork and eBusiness in Architecture, Engineering and Construction 2022 (2023)
The construction industry grapples with stagnant productivity and slow digitalization, marked by data heterogeneity and silos. To tackle this, tech companies are leveraging Artificial Intelligence (AI) to analyze complex data and enhance decision-making. This study assesses 236 AI-focused tech companies in Architecture, Engineering, and Construction (AEC), using 16 variables. Their impact, measured by social network followers and revenue, revealed correlations with location, company characteristics, and technology type. The findings indicate untapped innovation potential in unexplored areas, highlighting opportunities for AI adoption and advancement in the construction sector.
Synergies Between Lean Construction and Artificial Intelligence: AI Driven Continuous Improvement Process
30th Annual Conference of the International Group for Lean Construction, 2022 (engl.)
Both, Lean Construction (LC) techniques and Artificial Intelligence (AI) methods strive for the continuous improvement of production systems in projects and organizations. A combined implementation of both approaches is an ongoing research area. Therefore, the question arises as to whether the added value generated by implementing both approaches jointly is greater than the added value generated by implementing them independently and what is the significance of people in their combined use. This paper explores theoretically the potential of synergies between LC and AI in the AEC sector with exemplary use cases as well as their resulting effects.
An Empirical Evaluation of Predicted Outcomes as Explanations in Human-AI Decision-Making
Proceedings of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2022 (engl.)
In this work, we empirically examine human-AI decisionmaking in the presence of explanations based on predicted outcomes. This type of explanation provides a human decision-maker with expected consequences for each decision alternative at inference time—where the predicted outcomes are typically measured in a problem-specific unit (eg, profit in US dollars). We conducted a pilot study in the context of peer-to-peer lending to assess the effects of providing predicted outcomes as explanations to lay study participants.
Designing a Human-in-the-Loop System for Object Detection in Floor Plans
Proceedings of the AAAI Conference on Artificial Intelligence, 2022 (engl.)
In recent years, companies in the Architecture, Engineering, and Construction (AEC) industry have started exploring how artificial intelligence (AI) can reduce time-consuming and repetitive tasks. One use case that can benefit from the adoption of AI is the determination of quantities in floor plans. This information is required for several planning and construction steps. Currently, the task requires companies to invest a significant amount of manual effort. Either digital floor plans are not available for existing buildings, or the formats cannot be processed due to lack of standardization. In this paper, we therefore propose a human-in-the-loop approach for the detection and classification of symbols in floor plans.
Predicting the construction duration in the predesign phase with decision trees
In ECPPM 2022-eWork and eBusiness in Architecture, Engineering and Construction 2022
This paper assesses decision trees’ (DTs) effectiveness in predicting construction duration during the predesign phase. Surveying the German AEC industry for expected prediction accuracy, the authors compare five DTs (Random Forest, GBR, XGBoost, LightGBM, CatBoost) with artificial neural networks (ANN) and linear regression models (LRMs) using two residential project data sets. Performance indicators (MAE and MAPE) demonstrate that DTs outperform ANNs and LRMs. Achieving a 26% expected prediction accuracy, data set 1 yields a 13.48% MAPE, while data set 2 approaches it with a 26.45% MAPE. This underscores DTs’ practical potential amid increasing construction data generation, emphasizing the need for further testing their explainability in predicting construction duration.
Artificial Intelligence for the Construction Industry - A Statistical Descriptive Analysis of Drivers and Barriers
Part of the Advances in Intelligent Systems and Computing book series (AISC,volume 1410) (2021)
AI applications in construction are underutilized despite the transformative impact in other industries through digitalization and Industry 4.0. This study aims to identify drivers and barriers for AI implementation in construction and related sectors. Analyzing data from over 100 sources, including 35 selected for relevance, involving feedback from 36,269 companies in Europe and North America, revealed 23 factors. Among them, 15 were drivers (e.g., productivity enhancement), and 21 were barriers (e.g., skilled labor shortage). Economic and technological factors predominantly influenced AI adoption, while social factors featured more in barriers, with fewer political considerations.
Konzept einer durchgängigen Informationskette mit Methoden der Künstlichen Intelligenz am Beispiel der Lieferkette von Beton
Proceedings of the 6.th International BBB-Congress, 2021
Bauprojekte und deren Lieferketten sind durch die Vielzahl an Schnittstellen komplex. Informationsketten weisen viele Medienbrüche auf und ein Informationsaustausch findet häufig analog statt. Auf der anderen Seite werden durchgängige Informationsketten für einen stabilen und verschwendungsarmen Bauablauf benötigt. Im Rahmen dieser Veröffentlichung wird ein zweistufiges Implementierungsmodell vorgestellt. Das Modell zielt darauf ab, eine durchgängige Informationskette mit reduzierter Ressourcenverschwendung zu schaffen. Dabei wird das Modell beispielhaft auf die Wertschöpfungs- und Lieferkette Betonbau angewendet. Hierbei wird in der ersten Stufe auf Methoden der Künstlichen Intelligenz zurückgegriffen und in der zweiten Stufe auf eine Standardisierungsmaßnahme.
Digitalization of Lean Learning Simulations: Teaching Lean Principles and Last Planner® System
29th Annual Conference of the International Group for Lean Construction, 2021 (engl.)
In dieser Veröffentlichung werden Lean-Simulationen auf einer digitalen Plattform analysiert, die Echtzeit-Interaktionen zwischen mehreren Teilnehmern ermöglicht. Zwei Simulationen wurden in einem dreistufigen Entwicklungsprozess erstellt, die sich auf Lean-Prinzipien und das Last Planner®-System konzentrieren. Teilnehmerbewertungen durch Fragebögen ergaben durchweg positive Ergebnisse, wobei Technologie, Spaß und Design hoch bewertet wurden. Die Ergebnisse belegen die erfolgreiche Interaktion über die gewählte digitale Plattform und unterstreichen die Flexibilität, einfache Integration, niedrige Kosten und hohe Nachhaltigkeit von digitalen Simulationen.
Human-AI Complementarity in Hybrid Intelligence Systems: A Structured Literature Review
28th Pacific Asia Conference on Information Systems , PACIS 2021 (engl.)
Hybride Intelligenz ist ein neues Konzept, das die Komplementarität von menschlicher Intelligenz und künstlicher Intelligenz (KI) betont. Eine wichtige Voraussetzung für die Zusammenarbeit zwischen Mensch und KI ist die Interpretierbarkeit der von der KI getroffenen Entscheidungen. Im Rahmen einer strukturierten Literaturrecherche identifizieren wir relevante Faktoren, die die Zusammenarbeit zwischen Mensch und KI beeinflussen.
Artificial Intelligence for the Construction Industry - a Statistical Descriptive Analysis of Drivers and Barriers
1st International Conference on Disruptive Technologies Tech Ethics and Artificial Intelligence , 2021 (engl.)
In the construction industry, applications of artificial intelligence (AI) are still uncommon. Digitalization, industry 4.0 and the increasing implementation of AI systems already show a rise in efficiency and innovation in many industries. Thus, there are still untapped opportunities for construction.
The aim of this work is to identify drivers and barriers for the implementation of AI in the construction sector and its adjacent industries. Data on pre-vailing factors was collected in a systematic review of more than 100 studies, surveys, and statistics…
BIM Basics: BIM und KI in Wissenschaft & Unternehmenspraxis
buildingSMART Deutschland, 2021 (de.)
Die Rahmenbedingungen verändern sich im Zeichen des Wandels und der stetig notwendigen Erneuerung für Mensch, Gesellschaft und Unternehmen gleichermaßen rasant. Die Digitalisierung berührt alle Lebensbereiche und ist zugleich Verheißung, Fluch und letztendlich unumkehrbar.
Auch wenn vieles heute ungeklärt erscheint und sich im Stadium der Erforschung der Zusammenhänge befindet, so verbindet diese Publikation die Stichwörter BIM und KI, digitale Geschäftsmodelle, IoT und E-Commerce und zeigt Beispiele aus der Unternehmenspraxis auf.
Specification of a standardized format to define and exchange Model View Definitions with Exchange Requirements and Validation Rules.
Technical Services of buildingSMART International Ltd. , 2021 (engl.)
Technische Dokumentation auf GITHUB: neue Version des buildingSMART Standards mvdXML in Version 1.2. Dabei handelt es sich im Moment um eine Draft Version.
Ressourceneffizientes Bauen mit Beton unter besonderer Berücksichtigung der Methode der Lean Construction
Ressourceneffizienter Beton-Zukunftsstrategien für Baustoffe und Baupraxis (2020)
Das Bauwesen ist ein bedeutender Wirtschaftssektor, der 13% des weltweiten Bruttoinlandsprodukts ausmacht. Trotz dieser Wirtschaftsrelevanz bleibt die jährliche Produktivitätssteigerung in den letzten 20 Jahren mit etwa einem Prozent begrenzt. Lösungsansätze für dieses Dilemma liegen in den Grundprinzipien des Lean Thinkings, definiert als Kundenwert, Identifikation des Wertstroms, flussorientierte Produktion, bedarfsorientierte Fertigung und Streben nach Perfektion. Im Bauwesen, als Lean Construction bekannt, umfasst dies die Gestaltung von Produktionssystemen, kooperative Arbeitsplanung und integrierte Projektabwicklung. Im Hinblick auf das ressourceneffiziente Bauen mit Beton wird hier im Speziellen auf die Gestaltung und Steuerung von Produktionssystemen eingegangen.
A requirement model for Lean leadership in construction projects
Proceedings 28th Annual Conference of the International Group for Lean Construction (IGLC) , 2020 (engl.)
A growing number of construction companies are embracing Lean philosophy to enhance customer value and identify waste. A recent survey of Lean Construction Institute members highlighted managerial barriers as primary obstacles. Many companies prioritize Lean methods/tools, often neglecting social factors crucial for cultural change. Site managers, acting as links between value-adding personnel and top management, play a crucial role. A necessary shift from a managerial to a leadership role demands supporting subcontractor collaboration and continuous individual skill improvement. Drawing from comprehensive literature, the authors propose a Lean leadership requirement model for construction managers, offering a foundation for further research and training programs.
Building the future of the construction industry through artificial intelligence and platform thinking
Digitale Welt, 24.01.2020 (engl.)
Data in the construction industry is heterogeneous, organizations do not work closely together, and construction software is highly specialized for individual users and applications. As a result, knowledge from previous construction projects is often not shared, linked, or transferred to subsequent projects. Additionally, the manual work on-site leads to long and unstable design and construction processes. Based on a review of common challenges in this work a new vision for the Architecture, Engineering, and Construction (AEC) Industry is developed...
Mehr Sicherheit dank Künstlicher Intelligenz
BauPortal 4|2020 (de.)
Das Forschungsprojekt „Smart Design and Construction“ (SDaC) […] will die Sicherheit auf Baustellen mithilfe Künstlicher Intelligenz optimieren. In verschiedenen Anwendungsfällen werden u. a. das teilautomatisierte Erstellen von Gefährdungsbeurteil-ungen sowie das teilautomatisierte Erkennen von Gefahren auf Bestellen durchlaufen.
Wie die Applikation gestaltet sein könnte, erfahren Sie hier:
The General Contractor Response to Platform Ecosystems
IGLC, 27.07.2019 (engl.)
Platforms enable value-creating interactions between producers and customers by mediating between their users. Supported by digitization, platforms use large datasets and integrated production systems to enhance the customer and producer experience. The platform's business model is expanding in the economy as digitization increases...
How will a future platform disrupt construction industry?
English 
German