Utilizing BIM and Neo4j to prioritize clash resolution and enhance efficiency in construction planning

Group Research Project

Partner: Hairuo Zhao
Advisor: Yuqing Hu
2023

Nowadays, the clash detection technique is frequently used in the construction industry, especially for complex projects. Clash happens frequently in construction projects, which prevents building elements from fitting or functioning correctly, and this will cause further impact on cost, schedule, and quality control.

The project aims to leverage BIM (Building Information Modeling) in conjunction with graph theory to optimize the clash reviewing process.

Considering building as an interdependent system, conflicts are resolved by focusing on the dependencies between the components.

The goal of this project is to improve the clash solution by optimizing the clash correction sequence.

Currently, clash detection is typically done using Navisworks by creating search sets and reviewing two building systems at a time to identify potential clashes between them.
One significant limitation in the current process is that BIM applications can only define clashes individually, without any grouping or categorization. This limitation becomes particularly challenging because there are often numerous clashes to review, making it a time consuming task for BIM/VDC engineers.

Motivation and Methodology

Through graph analysis, it helps to visualize the relations between the building components which complements the clash detection report to ease the coordination process. It is able to identify and understand the complexities and potential conflicts in the construction project. Integrating graph analysis algorithms, such as centrality, community detection, and similarity analysis, allows the project team to plan the work more strategically and orderly, reducing the time and resources spent on identifying, sorting, and resolving clashes. This leads to a smoother coordination process, ultimately ensuring a more efficient and cost-effective project construction process.

Algorithm and Data Coordination

The logic of solving topological order problems aligns with our problem statement and research approach, which focuses on addressing clash resolutions. This order is frequently used in solving construction sequencing and task dependence problems. A common algorithm design pattern is listed as:

1. Figure out how to get the first thing.

2. Remove the first thing from the problem.

3. Repeat until the problem is solved.

The topological graph is effective in identifying the priority of each involved building components through their centrality and degree shown in the graph. This not only highlights the sequential relationships among the nodes but also offers a clear and structured representation of the resolution sequence for each component involved in clashes.

Tools and Strategies

The analyses of graph centrality, similarity, and community detection are conducted to identify repetitive components and clash types in the overall clash detection, aiming to enhance the thoroughness of resolution strategies. Centrality analysis is used to organize the importance and urgency of the building component relocation or adjustment. Considering the potential recurring clash types across various project areas, engineers could conduct similarity analysis to indicate the solutions for similar clash types, further increasing the overall detection efficiency. The similarity in relationships will be determined by how closely two connected nodes resemble each other. The same method can be applied for the clash resolution if two linked nodes are similar.

Moreover, the displacement of building components could impact the repositioning of previously resolved clashes. Employing cycle detection is crucial in minimizing redundant work during component relocation. It also assists engineers in determining the optimal sequence for implementing solutions.

A total of three rounds of clash detection are performed. The clash detection setting is detecting hard clashes only with a tolerance of ¼ inches to account for small installation deviations and potential materials thermal expansions. The following chart shows the number of clash detected in each round of detection using Navisworks.

The node properties include their item IDs being GlobalId, name, and type of the component for easier and more accurate identification of the component. To ensure a clear and concise relationship graph for future analysis, repetitive nodes from various clashes are eliminated.

Three building system models are provided for this research, each of them are: Structure, Architecture, and Mechanical, Electrical, Plumbing, and Fire Protection (MEPF). They are paired in twos to detect clashes in Navisworks.

Outcome and Expected Contributions

Three building system models are provided for this research, each of them are: Structure, Architecture, and Mechanical, Electrical, Plumbing, and Fire Protection (MEPF). They are paired in twos to detect clashes in Navisworks. A total of three rounds of clash detection are performed.

The degree centrality of nodes and edges is the key factor for clash priority identification. Typically, building components with the highest clash occurrences need more consideration and higher priority for resolution as they impact more building components. The centrality degrees are sorted from the highest to the lowest. The centrality degrees are the representative of the number of clashing building components to a specific element, which would be a significant factor to be considered during the determination of resolution sequence. With the application of Neo4j, the degree centrality can be reflected as a node property which will provide more visible and accessible information to VDC engineers for their future work and analysis.

The connected component analysis identifies the discrete clusters within the model, and it is crucial for recognizing potentially interacting groups. Triangle detection presents areas where three components are mutually interconnected, which indicates a higher chance of complex clash situations and repetitive work. In the graph, the areas that show a dense network of pipes and joists may require more intensive coordination because one change can possibly affect the whole system.

This research approach aims to provide a supporting visualization of clash priorities and help VDC engineers establish sequential relationships between clashes occurring in construction. With the implementation of Neo4j and Navisworks, aligned with the priority principles learned from professionals, the graph analysis outlined in this study provides a comprehensive clash report. Given both the clash report and the visualization of the sequential relationship between the building components, VDC engineers should be able to identify each clash occurring during the construction process and will be able to know any adjustments in the location or shape of subsequent components. This approach aims to reduce the manual checking efforts in order to streamline the workflow.