Intelligent rebar layout in RC building frames using artificial potential field

https://doi.org/10.1016/j.autcon.2020.103172Get rights and content

Highlights

  • Develop an artificial potential field framework for intelligent rebar layout.

  • Automatically address rebar collision and congestion in RC beam-column joints.

  • Model rebars as intelligent agents with the ability of avoiding obstacles.

  • Propose an improved artificial potential field method.

  • Automatically generate BIM for 3D visualization of rebars.

Abstract

Reinforced concrete (RC) beam-column joints are the critical components of RC building frames. Congestion and even collision of reinforcing bars (rebars) encounter frequently in the core region of a joint. The building information modeling (BIM) technology is helpful to detect the rebar collision, which has to be revised manually. In this paper, inspired by collision-free capability of intelligent agent, an automatic rebar layout framework based on artificial potential field (APF) is proposed to solve the issues of rebar collision and congestion during the design stage. The rebars in a beam-column joint are regarded as agents, and a novel APF method is proposed to guide the moving trajectory of each agent. The APF method ensures that each rebar-agent can reach its goal point and escape from the trapping of local-minimum point. An Autodesk Revit add-in is developed to save modeling time and improve accuracy. Experiments on a two-story RC frame validate the applicability and efficiency of the intelligent rebar layout framework. It can be shown that the developed framework can avoid the rebar collision and congestion at beam-column joints quickly and effectively.

Introduction

The beam-column joints are the critical components of a reinforced concrete (RC) building [1]. To strengthen the building structure, a larger number of rebars must be laid out at the joints and an accurate layout of rebars must be ensured [2,3]. Rebars in the joint are often intersected or even overlapped in three directions [4], thus likely causing the collision or congestion among them. In designing beam-column joints, rebar collisions and congestions must be identified and avoided to ensure the structural safety and constructability. Rebar collision implies the spatial overlapping between rebars from different components, while rebar congestion means that the clear distance between adjacent rebars or the clear distance between the rebar and concrete surface is too small to meet the construction requirements. At the construction site, workers usually can lay out rebars according to as-design drawings without much difficulty if no rebar collision and congestion exist. When encountering with a rebar collision or congestion issue, the engineer and worker have to identify it, and adopt an expedient procedure to resolve it manually. If the re-location of rebars is inappropriate, the seismic performance and capacity of components will be decreased. Compared with the manual approach, an automatic layout would be less tedious and more cost-saving and will have a positive impact on construction quality. For this purpose, an intelligent rebar layout able to avoid rebar collision and congestion automatically was studied.

In the architecture, engineering, and construction (AEC) industry, rebar design (e.g., number and diameter of rebars) is usually done with the use of the computer software [5]. For the rebar layout, an automatic collision-free design method is generally unavailable in practice. Some firms did notice to integrate intelligent design with building information modeling (BIM) technology in their projects [6]. A BIM model is a project simulation system consisting of a three-dimensional (3-D) model representing the project components linked to all required project information pertaining to planning, design, construction, and operation [7]. Engineers may use the 3-D BIM model to detect the possible rebar collision visually, but the process of rebar modeling is time consuming and the visual detection would lead to mistakes. Current computer software can automatically detect rebar collision through the use of application programming interface (API), but the rebar layout still need be adjusted manually to ensure collision free. Moreover, in an RC building frame, there are many beams and columns requiring specific rebars. It is very time-consuming, dull, and repetitive to build a BIM model. An automatic generation tool for BIM models is essential to advance the designing efficiency of RC structures.

In reviewing the literature, several researchers have studied the collision detection problem of construction. Han et al. [8] proposed a piping BIM model and utilized Autodesk Navisworks [9] software to pinpoint potential collision locations and generate a collision detection report. The colliding piping was then adjusted with a few times of correction manually. Lai et al. [10] developed a simple and efficient collision detection method to support the rendering of a virtual construction scenario in real time. Radke et al. [11] proposed an approach to solve the clash identification and resolution for mechanical, electrical, and plumbing (MEP) systems, based on the original design parameters of objects. Wang et al. [12] proposed a formalized scheme for the spatial clash coordination of MEP systems to the store clash analysis, resolution and management information. These methods mainly applied to MEP systems are not able to provide the automatic collision solution of rebars in RC joints.

There is few work done on rebar layout. Navon et al. [13] proposed a diagnosis module for diagnosing and correcting rebar constructability issues. Rebar congestion, rebar collision, concrete cover, and reinforcement radio were tested based on the rebar constructability knowledge. Park [14] proposed a BIM-based simulator to perform clash detection and determine the placement sequence of rebars. A clash was automatically identified if the distance between rebar centerlines was less than the rebar diameter. However, these aforementioned studies were dedicated to the spatial rebar collision detection rather than a collision-free layout. Mangal et al. [15] proposed a framework of rebar clash resolution for RC beam-column joints based on the BIM and genetic algorithm (GA), where clashes were identified when the total number of rebars exceeded the rebar number range. However, only a limited number of rebars was involved to detect the rebar collision or congestion, and the spatial geometry of rebars is seldom considered in their study. The aforementioned studies did not automatically provide alternative solution to adjust the placement of colliding rebars.

In this paper, an intelligent rebar layout framework is proposed to complete the rebar collision-free design of a 3-D beam-column joint automatically. Inspired by the collision-free capability of robot or intelligent agent, each rebar is regarded as an agent that intelligently moves in a complex structure and the moving path of agent is then used to guide the layout of rebar. Currently, the common path planning methods for robotics to complete the obstacle avoidance include artificial potential field (APF) [16,17], reinforcement learning (RL) [18], and heuristic algorithms (e.g., Firefly Algorithm (FA) [19] and Particle Swarm Optimization (PSO) [20]). In comparison with other methods, the APF method has some promising features, such as conceptual simplicity, computational efficiency, and clear mathematical explanation. The APF can provide consistent and reliable paths of the agent in a fast way. Therefore, the APF method is more suitable for the framework of heavy and tight rebar layout tasks. In this study, the APF method is revised for the task of intelligent rebar layout, so that the rebar collision and congestion could be detected and avoided automatically. As the traditional APF method has the drawbacks of goal non-reachable problem with obstacle nearby (GNRON) and local-minimum trapping problem, we proposed a novel APF, in which the repelling potential function of obstacles is modified and the repulsive force from obstacles is revised heuristically. Furthermore, the BIM model of a beam-column joint was generated automatically using the BIM software Autodesk Revit [21] API to represent the rebar layout detail. Therefore, this paper aims to propose an intelligent rebar layout framework based on a multi-agent system to generate rebar layout without the rebar collision and congestion. The novel APF method is applied to complete path planning with obstacle avoidance. A Revit add-in is developed to achieve the automatic generation of BIM model to improve the efficiency of RC building frame design.

This paper is organized as follows. In Section 2, the problem of the rebar layout in the joint is described and the traditional APF method is stated. Section 3 proposes the intelligent rebar layout framework using the APF method. Section 4 describes a novel APF method to cope with the GNRON and local-minimum trapping. Section 5 introduces an approach to achieve the automatic generation of a BIM model using Autodesk Revit API. Section 6 presents a two-story RC building frame including six different types of RC beam-column joints as the illustrative example to show the applicability of the proposed framework. The paper is concluded in Section 7.

Section snippets

Rebar layout at the RC beam-column joint

The simple RC building frame shown in Fig. 1 consists of horizontal components (beams) and vertical components (columns). The beam-column joint is a zone where beams are overlapped and intersected with the column in all three directions. The rebar layout for each beam or column is relatively simple, but it is rather complex at the beam-column joint [22]. Rebars can be classified into two basic categories: longitudinal rebars and transverse rebars. The former ones are placed along the component

Intelligent rebar layout framework using the APF method

An intelligent rebar layout framework using the APF method is proposed to detect and solve rebar collision and congestion problems. According to the extracted design information from the structure design software, the RC beam-column joint can be modelled in a 3-D workspace. All longitudinal rebars in RC components can be laid out with no collision and congestion. Therefore, the proposed framework (Fig. 4) is divided into three phases: (1) Extraction of design information, (2) Building a rebar

A novel APF method for intelligent rebar layout framework

The APF method is elegant, reliable, and efficient in terms of path planning process [29]. Using the traditional APF method as described in Section 2.2, the rebar-agent can reach its goal point in most cases. However, according to the available rebar layout results obtained from the traditional APF method, in certain situations, the rebar-agent cannot reach its goal point and thus the task of rebar layout would not be completed. The reason is that the traditional APF method has the inherent

Automatic generation of a BIM model

BIM technology has been utilized by the AEC industry to enhance the quality and accuracy of design and construction of RC structures [37]. Potential constructability problems can be discovered and solved during the design stage using a 3-D BIM model. The design information of an RC building frame such as components' geometrics can be seamlessly exchanged between different 3-D applications through BIM. According to the design information and the simulated result from the proposed rebar layout

Example 1 - comparisons with other path planning methods

A simple example is considered to compare the APF method with other path planning methods such as RL [18] and PSO [19]. As shown in Fig. 12, a rebar-agent moved in a simple workspace, which size is 200 mm × 200 mm × 200 mm. A column rebar is placed at (100, 50, 0) and a beam rebar in X direction is placed at (0, 150, 100). The length of two rebars are both 200 mm and the diameter is 20 mm. The rebar-agent represents a rebar of 20 mm diameter (denoted as Φ 20 mm). Its start and goal points are

Conclusions

Rebar collision and congestion often occur at beam-column joints, but rarely considered during the design stage. Currently, workers must adopt an expedient procedure to resolve the rebar collision and congestion issues manually on construction site. This approach is tedious and time consuming with negative impacts on the construction quality. Using the BIM technology, each misplaced rebar requires a manual adjustment to ensure rebar collision and congestion free.

Therefore, an intelligent rebar

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was supported in part by the Graduate Scientific Research and Innovation Foundation of Chongqing, China (Project No. cstc2019yszx-jscxX0001) and the National Natural Science Foundation of China, China (Project No. 61803054).

References (40)

  • S. Rowlinson et al.

    Implementation of building information modeling in construction: a comparative case study

    American Institute of Physics Conference

    (2010)
  • S. Azhar et al.

    Building information modelling: now and beyond

    Construction Economics and Building

    (2015)
  • N. Han et al.

    Collision detection of building facility pipes and ducts based on BIM technology

    Adv. Mater. Res.

    (2012)
  • Anon, Autodesk Inc., Navisworks, California, Available from: https://www.autodesk.com/products/navisworks/overview....
  • A.M. Radke et al.

    An automated approach for identification and resolution of spatial clashes in building design

    IEEE International Conference on Industrial Engineering and Engineering Management

    (2009)
  • R. Navon et al.

    Automated rebar constructability diagnosis

    J. Constr. Eng. Manag.

    (2000)
  • U.Y. Park

    BIM-based simulator for rebar placement

    Journal of the Korea Institute of Building Construction

    (2012)
  • M. Mangal et al.

    Automated clash resolution of steel rebar in RC beam–column joints using BIM and GA

    International Symposium on Automation and Robotics in Construction

    (2017)
  • S.S. Ge et al.

    Dynamic motion planning for mobile robots using potential field method

    Auton. Robot.

    (2002)
  • T. Weerakoon et al.

    An artificial potential field based mobile robot navigation method to prevent from deadlock

    Journal of Artificial Intelligence and Soft Computing Research

    (2015)
  • Cited by (0)

    View full text