Sign Up
ENعربي
All articles

Project Management

Artificial Intelligence in Lean Construction Management (LCM)

  • Introduction

The construction industry is increasingly challenged by complex projects influenced by environmental, technical, and safety factors. This complexity spans all project phases, demanding effective management approaches. Lean Construction Management (LCM), which emerged in the early 2000s, offers a framework to manage this complexity by maximizing value, minimizing non-value-added activities, and eliminating waste. However, conventional LCM methods may not be sufficient to handle the complexities of modern construction processes. This has led to the exploration of advanced tools, particularly Artificial Intelligence (AI), to enhance LCM practices. This literature review aims to analyze existing studies on the application of AI in LCM, offering insights into how AI tools can support LCM principles and identifying future research directions.

  • Background Concepts

Lean Construction Management (LCM), derived from Lean Management, focuses on designing production systems that minimize waste of materials, time, and effort while maximizing value. LCM improves competitiveness and productivity by eliminating non-value-added activities and enhancing workflow processes. Key objectives of LCM include: 1) reducing non-value-adding activities and waste, 2) reducing lead time and variability, and 3) simplifying workflows, increasing flexibility, and improving transparency.

Artificial Intelligence (AI), initially defined as programs that learn from experience, has evolved to encompass computer-assisted systems that perform cognitive functions associated with humans. AI is divided into two main paradigms: Symbolic and Connectionist. Symbolic AI operates on symbolic expressions and rule-based reasoning, requiring human intervention and not learning from input data, such as in knowledge-based databases and expert systems. Connectionist AI, on the other hand, uses learning algorithms and neural networks to correlate input and output data without explicit rules and including tools such as machine learning, artificial neural networks, and deep learning.

III. Research Methodology

This review adopts a systematic literature review (SLR) methodology to ensure a rigorous and transparent analysis of existing literature. The SLR process involves five steps: planning, searching, screening, extraction and synthesis, and reporting. The planning phase used Cooper's (1988) taxonomy to scope the review. The focus was on the notions of LCM and AI; the goal was to analyze the nexus between them; the review was comprehensive; the organization was to develop a conceptual framework; and the target audience included scholars and practitioners in construction management.

The searching process used keywords related to LCM and AI combined with Boolean operators like "AND" and "OR". Keywords included "lean construction", "artificial intelligence", "robotics", "machine learning", and "optimization". The search included the Scopus, EBSCO, and ScienceDirect databases. The screening process involved removing duplicates from the search results and selecting relevant papers by carefully reading abstracts and then the full text. Initially, 246 de-duplicated articles were identified, which were then narrowed to 55 relevant articles through forward and backward tracking of citations. In the extraction and synthesis phase, AI tools were categorized as either symbolic or connectionist, and the LCM principles to which they were applied were coded. The reporting phase included a bibliometric analysis and content analysis to develop a classification scheme for AI tools in LCM.

  • Bibliometric Analysis

The bibliometric analysis focused on keyword co-occurrence, country-specific information, and historical trends of publications and citations. The keyword co-occurrence analysis identified 41 co-occurred terms, where the size of the circles is proportional to the number of times that a given keyword appeared in the title, abstract and authors’ keywords. The analysis revealed that while many terms related to construction and lean management were used, the terms "artificial intelligence" and "AI" were not frequently used. This is because "AI" is an umbrella term, and researchers often use specific AI-related terms such as "Building Information Modeling (BIM)", "simulation", "optimization" and "automation". The growth of keywords over time shows a significant increase since 2014, indicating a surge of interest in these fields in the past eight years.

The country-specific information showed that the United States, the United Kingdom, and China are the top three countries in the number of publications in the field of AI in LCM. The co-authorship network shows that the United States, the United Kingdom, and Australia are the most cooperative countries. A bibliographic coupling analysis reveals that the United States and the United Kingdom are the most central countries in the network, implying that their publications are highly cited in this field. The historical trend analysis showed an overall increased number of published papers over the 1998–2021 period, with a peak in 2015 and an average annual growth rate of 11.2%. The Reference Publication Year Spectroscopy (RPYS) method indicated a tendency to cite more recent references, reflecting the evolving nature of the field.

  • Research Findings and Discussion

The reviewed articles were categorized based on the two paradigms of AI: Symbolic and Connectionist. Symbolic AI tools such as simulation, fuzzy logic, expert systems, and visualization tools were frequently used to reduce waste and improve processes. Connectionist AI tools, mainly optimization tools and robotic technologies, were employed to manage uncertainties and improve on-site modular building.

The articles were further classified based on seven key principles of LCM for which AI tools have been utilized:

  • Waste Reduction: AI tools like Discrete Event Simulation, Artificial Neural Networks, Optimization, KanBIM, and Value Stream Mapping are used to identify, visualize, and manage construction wastes.
  • Reduction of Lead Time: AI tools such as Simulation Analysis, Extended Reality, Scheduling Optimization, Genetic Algorithms, Hybridized BIM, and Design Optimization are used to reduce processing, inspection, wait, and move times.
  • Reduction of Variability: Computerized Integrated Project Management Systems, BIM, Genetic Algorithms, Linked-Data Based Constraint-Checking, and Discrete Event Simulation are employed to improve the reliability of look-ahead planning and minimize variations in time, cost, workflows, and processes.
  • Enhancing Construction Safety: AI-based applications are utilized to determine hazard, exposure, and severity levels of construction safety risks, while BIM-based visualization frameworks are used for safety planning.
  • Increasing Productivity: The integration of AI tools such as robotics, image processing, Last Planner Systems, and computer-based simulations is used to enhance construction productivity.
  • Developing Project Delivery Systems: Integrated Project Delivery (IPD) and AI tools like BIM are used to support project delivery through enhanced collaboration, trust, and early stakeholder involvement.
  • Improving Construction Efficiency: Various AI tools, including simulation modeling packages, BIM, 3D modeling tools, and AI-based measurement software, are used to improve overall construction efficiency.

The analysis showed a significant increase in publications since 2014, attributed to the growing recognition of both LCM and AI in academia. The roadmap for future research, derived from the limitations and recommendations in the reviewed literature, includes:

  • The hybridization of AI tools and LCM methods, such as the integration of machine learning, cloud computing, optimization, and simulation into BIM.
  • Applications in real-life settings, extending AI applications beyond controlled environments to real-world construction projects and supply chains.
  • Compatibility of existing AI tools with existing LCM methods, addressing the functional limitations of tools like fuzzy logic, BIM, and simulation, and suggesting the development of domain-specific AI tools.
  • Big data analytics, leveraging large datasets in areas such as BIM, construction implementation, construction management, production, and real-time automation.
  • Conclusion

This review has highlighted the increasing role of AI in enhancing LCM practices in the construction industry. There is a gap in the literature regarding the direct nexus between LCM and AI. However, the review has provided a classification scheme for AI tools based on the key principles of LCM and proposed a roadmap for future research. The practical implications of this research include a better understanding of the link between AI and LCM for construction managers and a practical toolbox of AI tools to enhance LCM practices. The limitations of the research include its focus on the construction phase and the need for further empirical studies. Future research should extend to all phases of construction projects and focus on the development of the next generation of AI tools to facilitate the implementation of LCM.

(Dumrak & Zarghami, 2023)

Dumrak, J., & Zarghami, S. A. (2023). The role of artificial intelligence in lean construction management. In Engineering, Construction and Architectural Management. Emerald Publishing. https://doi.org/10.1108/ECAM-02-2022-0153

  • What is Lean Construction Management (LCM) and why is it important?

LCM originates from Lean Management and aims to improve the construction industry's competitiveness and productivity by eliminating waste (non-value-added activities) and improving workflow processes. It focuses on maximizing value and minimizing waste throughout a construction project's lifecycle to prevent cost and schedule overruns, reduce waste of materials, time, and effort, and improve overall construction performance. While LCM offers several benefits, implementing it can be challenging due to the complexity of construction processes. This is why integrating methods from other disciplines, like AI, is often necessary to facilitate LCM implementation.

  • What role does Artificial Intelligence (AI) play in Lean Construction Management?

AI provides advanced tools to deal with the complexities of construction processes, supporting LCM by improving planning, coordination, and control, maximizing value, minimizing non-value-added activities, and eliminating waste. AI tools can enhance various principles of LCM, such as waste reduction, lead time reduction, and safety improvement. Although the application of AI in LCM is still in its early stages, AI is considered an effective tool that opens new pathways to achieve the key principles of LCM.

  • What are the two main paradigms of AI and how are they applied in LCM?

The two main paradigms are Symbolic AI and Connectionist AI.

  • Symbolic AI: Relies on explicitly coded rules and knowledge-based models. In LCM, Symbolic AI tools like simulation tools, fuzzy logic, expert systems, and visualization tools are used to reduce waste, improve processes, workflow, and productivity by generating rules for software to follow.
  • Connectionist AI: Uses learning algorithms and neural networks to correlate input and output data without codified expert knowledge. In LCM, Connectionist AI tools, including optimization tools and robotics, overcome multiple constraints, achieve optimal decisions, manage uncertainties, and assist with on-site modular building.
  • What are the key areas (principles of LCM) where AI tools are being applied?

AI tools are used in seven key areas related to the principles of LCM:

  • Waste Reduction
  • Reduction of Lead Time
  • Reduction of Variability
  • Increasing Productivity
  • Developing Project Delivery Systems
  • Enhancing Construction Safety
  • Improving Construction Efficiency
  • How is AI used to reduce waste in construction projects?

AI assists in effectively identifying, visualizing, and managing construction wastes. Tools such as Discrete Event Simulation, Artificial Neural Networks, Optimization algorithms, KanBIM, Bluetooth Low Energy, and Value Stream Mapping help maximize customer value through process optimization.

  • How does AI contribute to reducing lead time in construction?

AI tools are leveraged to ensure appropriate task handling, improve scheduling, and eliminate inefficiencies. Techniques include Simulation Analysis, Extended Reality, Scheduling Optimization, Genetic Algorithms, Hybridized BIM, and Design Optimization.

  • Which countries are leading in research related to AI and LCM?

The United States of America, the United Kingdom, and China are the top three countries in terms of the production of research articles, followed by Brazil and Israel. The United States, the United Kingdom, and Australia are the most cooperative countries in collaborative research.

  • What are the recommended future research directions for AI in LCM?

Four research directions are recommended:

  • Hybridization of AI Tools and LCM Methods: Integrating various AI tools like machine learning, cloud computing, optimization, simulation, and cognitive computing technologies into BIM.
  • Applications in Real-Life Settings: Extending AI applications beyond studied publications into real-world construction projects to overcome limitations of controlled environments, data constraints, and lack of supportive models.
  • Compatibility of Existing AI Tools: Developing domain-specific AI tools tailored to the specific characteristics of construction projects to address functional limitations of generic AI tools.
  • Big Data Analytics: Utilizing big data analytics in areas such as BIM, construction implementation, management, production, and automation to improve efficiency and reduce waste