As in manufacturing with its Industry 4.0 transformation, the enormous potential of artificial intelligence (AI) is also being recognized in the construction industry. Specifically, the equipment-intensive construction industry can benefit from using AI. AI applications can leverage the data recorded by the numerous sensors on machines and mirror them in a digital twin. Analyzing the digital twin can help optimize processes on the construction site and increase productivity. We present a case from special foundation engineering: the machine production of bored piles. We introduce a hierarchical classification for activity recognition and apply a hybrid deep learning model based on convolutional and recurrent neural networks. Then, based on the results from the activity detection, we use discrete-event simulation to predict construction progress. We highlight the difficulty of defining the appropriate modeling granularity. While activity detection requires equipment movement, simulation requires knowledge of the production flow. Therefore, we present a flow-based production model that can be captured in a modularized process catalog. Overall, this paper aims to illustrate modeling using digital-twin technologies to increase construction process improvement in practice.

In the industrial sector, products evolve significantly over their operational life. A key challenge has been maintaining precise, relevant engineering data. This paper explores the digital twin concept, merging engineering and operational data to enhance product information updates. It examines digital twin applications in construction, material flow, manufacturing and production, citing battery production and additive manufacturing. Digital twins aid in analyzing, experimenting with, and refining a system’s design and its operation, offering insights across product and system lifecycles. This includes tackling data management and model-data consistency challenges, as well as the recognition of synergies. This paper emphasizes sustainable, efficient management of engineering information, reflecting shifts in product longevity and documentation in industrial products and machinery.

This paper evaluates the applicability of an autonomous driving dumper for a construction project. Conducted as part of a joint lecture between Stanford University and the Technical University of Munich in 2024, this study provides students with hands-on experience in field assessments of construction robotics. The dumper, developed in partnership between the Technical University of Munich and the industry partner Wacker Neuson, is analyzed using the Robotics Evaluation Framework. Autonomous dumpers present a promising solution in the construction industry, offering improved safety, productivity, and the ability to address labor shortages, particularly on small and confined sites. The Robotics Evaluation Framework used is a systematic methodology for assessing feasibility and performance across safety, quality, schedule, and cost metrics. The results demonstrate that the autonomous dumper improves safety through better ergonomics, enables operators to focus on higher-value tasks, and enhances overall efficiency. While quality remains comparable to conventional dumpers, the autonomous system experiences 20% longer cycle times, presenting a scheduling challenge. Despite this, cost analysis reveals amortization after 1.43 years or 75 projects, highlighting its economic viability and potential for fleet scalability.

Building construction is a one-of-a kind production with complexity caused by interdependencies between logistical- and construction-operation processes. Especially the finishing phase involves numerous trades who install a wide variety of materials and have to share logistical resources. Traditional planning approaches based on analytical tools and experience gained from historical data fall short in providing the support needed to evaluate different logistical strategies. Therefore, the present work aims at developing a user-friendly simulation model that considers both material supply and waste removal through a third-party logistics partner (TPLP). The process interdependencies are investigated using data from a real hotel building project. A kitting solution is evaluated as an alternative, using consolidation centers to reduce material handling on site. Comparing the supply- and reverse logistics of these alternatives, kitting appears to be more costly however the model does not capture potential improvements in performance of construction operations and coordination of processes.

The Work Density Method (WDM) is used in takt planning for defining zones with equal workload. To date, this method has been applied mainly to building construction. This paper investigates the WDM's applicability to equipment-driven processes in heavy civil engineering, specifically to the in-situ production of foundation piles for a highway infrastructure project. Two existing computer-based programs that support the application of the WDM, WoLZo and ViWoLZo, were used to find a suitable grid size based on data from a real-world project. The results show the potential of using the WDM (1) to define zones with equal workloads, given that pile groups are irregularly distributed over the construction site space, (2) to compare different scenarios based on work density as a metric (e.g., scenarios with different uses and sequencing of equipment), and (3) to derive a takt time and process duration when using multiple pieces of equipment that must coordinate their efforts and work in sync. Compared to the building construction application, the heavy civil engineering application reveals new requirements when using the WDM and takt planning in general, regarding the geometrical and logistical needs of equipment-driven operations that constrain how zones can be defined.