Research on Construction Performance Evaluation of Robot in Wooden Structure Building Method

Abstract

Robots have been increasingly involved in global construction and are considered one of the most promising solutions for the reform of the construction industry. The superiority of the robotic construction process compared to the present stage of actual construction with predominantly human participation is mainly reflected in the high efficiency and high accuracy of its construction, thus improving the benefits to the natural, economic, and social environment and significantly changing the current construction labor market. At present, the development and application of robotic construction technology worldwide are mostly at the laboratory stage and are not applied to actual construction projects on a large scale. In this paper, we explore the methods and steps of robotic involvement in the construction of a complete timber building through the reconstruction of a completed timber building, from the design process and construction process to the final evaluation of construction efficiency. Finally, by comparing the advantages and disadvantages of the two design and construction methods, manual construction and robotic construction, the problems and difficulties in current robotic construction and the opportunities and challenges faced by robotic automation in the construction industry are identified.

1. Introduction

1.1. Application Status of Robotics in Architecture

Advanced digital and mechatronic technologies have a profound impact on industries that require manufacturing processes to be performed on a large scale. Meanwhile, the combination of digital technology and robotics is now revolutionizing manufacturing and is considered one of the most promising solutions to reform the construction industry and will enable greater profitability, efficiency, safety, and security [1,2,3]. Yet, the construction industry, especially the field of building construction, has made very little profit from it. The low level of digitalization in the construction industry makes it difficult to solve the problems faced by the construction industry using digital methods. In addition. the use of digitalization and robotics in the construction industry is still limited and has many challenges [1,4]. In this paper, we explore the reasons for these deficiencies and how this could change in the future through [5]. The construction industry differs from mass manufacturing in that it is an activity associated with the creation of physical artifacts. Mass manufacturing is a product designed for mass production. In contrast, building construction products are large and unique in form, and they may be produced on site in temporary disorder and chaos, with workers on site. Therefore, the use of construction robots is still in its infancy in terms of single-point development and small batch trials in the production of components and on-site construction and has not yet been achieved on a large scale [6].

For construction, however, robotic fabrication can improve the energy environment by minimizing material waste and reducing construction waste; on the other hand, it can improve the economic environment by increasing efficiency, optimizing the use of labor, and improving working conditions. Therefore, exploring digital and robotic construction in the first place is of great importance. The combination of digital design and manufacturing processes allows to adjust the design and processes of a structure under construction in real time to optimize the performance and aesthetics of the structure. We see digital building fabrication as both a major challenge and a tremendous opportunity. In the field of robotic construction, the research has predominantly focused on the prefabrication stage of building component factories [7], and those that can be applied to the construction site are mainly concentrated in the field of partial construction, such as masonry brick walls, pouring concrete, etc. The main types in the laboratory research stage are mainly for the construction of complex geometric form structures of different materials [2,8,9,10], 3D printing concrete residential experiments [11], etc. This paper supports the discussion with an example of robotic fabrication of a complete timber building to explore the methods and steps of robotic involvement in the construction of a complete timber building, from the design process and construction process to the final evaluation of construction efficiency. In general, robotics fabrication is a multidisciplinary process that combines comprehensive challenges.

1.2. Advantages of Robotic Construction

This paper focuses on the digital process of building fabrication and the application of robotics in construction fabrication, based on the premise that it has significant advantages over traditional building fabrication in terms of natural, economic, and social environments.

The first is the role of robotic construction in the improvement of the natural environment. The building industry needs to develop highly efficient and energy-saving construction systems against the background of global natural environmental degradation [12]. Automated and parametric construction systems are capable of performing repetitive tasks with high accuracy and efficiency and can improve environmental benefits compared to traditional construction methods.

The construction industry is a highly dynamic sector with a high proportion of global energy consumption and emissions [13,14]. The efficiency of the robots means less delay and time savings. As a result, the high efficiency of the entire construction process also means less time spent using heavily polluting machines, thus contributing to a reduction in environmental pollution [14].

Meanwhile, the high-precision nature of the construction robots allows them to make fewer mistakes, thus saving time and resources and reducing waste. By consuming fewer resources and having shorter machine run times, the building construction process reduces its environmental impact [14].

Robotic construction also contributes to the improvement of the economic environment. Increased construction efficiency and reduced construction costs contribute to long-term economic values, and these are significant drivers for research and utilization of robotics [15].

The traditional construction industry is facing the challenge of being large but not excellent. Especially under the influence of the Coronavirus disease of 2019, the traditional construction methods of the construction sector have been heavily hit, and the rough development modes have proven unsustainable. The utilization of construction robotics can increase the economic value in the long term, one of the most important aspects of which is reflected in the ability to change the construction labor market. Robotic construction can largely minimize the number of builders and the contact between them, and such a construction model has obvious advantages in the global epidemic.

Meanwhile, continuous training in irreplaceable skills is essential, which can attract workers to highly skilled technologies to gradually shift the workforce from heavy manual labor to light physical or mental labor [16].

For improving the social environment, robotic construction replaces workers with robots to perform hazardous and heavy labor tasks, ensuring worker safety and creating a better working environment [14]. For example, heavy and low-skilled repetitive tasks such as bricklaying, which can cause back strain and other injuries to workers, can be better replaced by robots. At the same time, robotic construction sites are relatively safe and environmentally friendly, minimizing disruption to the surrounding neighborhood.

Historically, the construction industry has been inefficient and harmful to the environment, but robots are changing that. The impact of robotics on environmental, economic, and social sustainability is multifaceted. The adoption of robotics can reduce environmental impacts and improve resource efficiency, long-term economic value, productivity, quality, worker health, industry, and public well-being [14].

2. Literature Review

2.1. Robotic Construction Research

Although rare in the status of affairs, robotics is being requested by the construction industry [7]. The robot developed in recent years for application in on-site bricklaying builds walls six times faster than humans [17]. There is also a case study of assembly implementations on construction job sites that precisely and automatically locates the best picking location for a robotic arm through a static performance optimization process during the design phase of a construction project [10]. Another case study demonstrates how architects can use robotic involvement to complete the design and construction of complex wood structures. By utilizing the adaptive geometry system, the design and construction of the Timber Structure Enterprise Pavilion at the Horticultural Expo were accomplished with a high degree of precision [18].

Robots were also involved in the house-building phase. The proposed advanced technology consists of creating a house with complex walls with 3D-printed materials using a mobile and poly articulated robot [11,19]. At present, research teams such as the school of architecture of Tsinghua University in China and the University of Nantes in France have studied and entered the stage of industrialization promotion [20].

In terms of materials, at present, the experimental projects of robot construction are mostly bricks, concrete, and wooden timbers. Among them, wood is one of the most sustainable and renewable materials and is considered to be potentially the most suitable material for the digital age [21]. The research on using timber as a construction material mostly focuses on wood structure construction experiments [9,18,22], engineered timber parametric connection experiments, etc. [8,21]. However, there is little research on the whole-house construction direction of wood structure construction robots. This paper takes the robotic construction of a simulation and its validation of an actual experiment of a timber building, focusing on its parametric construction process and verifying the implementation ability and superiority of robotic construction.

2.2. Research Purpose

The following research addresses three goals. First and most importantly, it addresses the design of digital construction parameters for the complete construction steps. In addition, it optimizes the robot arm movement mode to avoid collisions at the construction site, so that the robot can independently construct the wooden building, and verifies the applicability and feasibility of robot construction from the experiment. Secondly, for buildings of the same design and size, the construction efficiency and construction accuracy of robot construction and manual construction are compared to verify the superiority of robot digital construction over traditional construction modes. Finally, this experiment explores the assembly strategy for the robotic on-site fabrication of a timber building.

3. Methodology and Material

Based on the framework given above, this section describes in detail the parameterization process and construction process of robot construction of a wooden building, as shown in Figure 1. Through the reconstruction project of a wooden house, the following process is mainly used to show the ability and advantages of robot construction. Through target building analysis (Section 3.1), robot hardware equipment introduction (Section 3.2), building material system introduction (Section 3.3), building composition analysis (Section 4.1), robot construction process analysis (Section 4.2), and construction evaluation (Section 5.1, it is ensured to give full play to the robot’s initiative and complete the wood structure building construction experiment.

3.1. Research Objective Building

The building to be constructed is a wooden building that has been designed and built by manual laborers. This paper uses a construction robot to reconstruct the target building, and the focus is to compare the advantages and disadvantages of manual and robotic construction for the same wooden building in terms of software modeling, construction efficiency, accuracy, and difficulty.

In addition, this experiment is limited by the laboratory robot type, which cannot reconstruct the target building on a 1:1 equal scale, so the experiment in this article is divided into a simulated experiment part and an actual experiment part. A simulation experiment is the basis of the study, which explores the whole process of implementing robotic construction in an ideal state, aiming to explore the digital process of the integrated automation method in building construction and to compare the simulation results of robotic construction with manual construction in terms of construction efficiency and construction difficulty. The actual experiment is a verification of the simulation experiment, aiming to verify the operational feasibility and stability of this digital process in the actual experiment.

Here, the simulation experiment uses a KUKA robot with an arm operating radius of 3.9 m to reconstruct the target building on an equal scale; the actual experiment is a laboratory-scale case study, using a KUKA robot with an arm operating radius of 0.9 m available in the laboratory to reconstruct the target building on a 1:5 scale.

3.2. Experimental Robotic Set-Up

To implement the various steps of timber frame construction, the construction robot is equipped with an end effector that includes a gripper for pick-and-place procedures and a nail gun for executing drive-in-nail routines, as shown in Figure 2.

3.3. Building Material System

The building material system for the wooden building consists of Japanese cedar. Japanese cedar is considered the most economically significant tree planted in Japan, and it is estimated that cedar accounts for approximately 40% of all timber planted in Japan [23]. As a major renewable resource for construction, the timber was chosen as one of the optimal building materials for several reasons. First of all, it is relatively cheap compared to other building materials. Second, wood possesses excellent durability and strength in comparison to its weight. Third is its flexibility to be processed into different shapes, which makes it an adaptable material. Fourth, the sustainability of wood is reflected in the fact that it is a locally available building material and a growing resource [9].

There are several problems in the field of wood construction research. Timber fabrication is mainly limited to mechanical processing of individual components and manual assembly. This method is highly resource intensive. Here, the advantages of robotic construction are obvious, as not only does its usage save a lot of time, but the ability to transfer parametric design data directly into the assembly operation allows the construction of non-standard timber structures to be fully automated.

4. Design and Implementation of the Experiment

4.1. Overview of the Building

4.1.1. Design Objective and Generative Tools

The target building is a wooden-framed two-story building with a floor area of 32.1 m² and a building height of 5.9 m, as shown in Figure 3. The building uses 105 mm timber, which is the most common size of cedar lumber used for construction in Japan, and the standardized size improves the constructability of designs and studies. The floor plan is set in a simple form and the modules are set in multiples of 105 mm timber (105 × 9 = 945 mm) for the overall design.

The building’s design is stored within the fabrication data structure. This data structure is built upon a graph, within which the nodes of the graph represent the individual timbers, including their parameters, such as position, rotation, neighborhood, and the built state. In addition, the graph allows the discrete assembly sequences to be computed from it [24,25].

4.1.2. Fabrication Sequence for Robotic Fabrication

In this construction method, the plan is divided into seven units. As shown in Figure 4, the building is divided from north to south into seven units, A, B, C, D, E, F, and G, which are perpendicular to the ground. To facilitate construction, each unit was set up with 6 to 9 layers. After completing the construction of each unit, a crane was eventually used to erect the units. Since the joints were formed as each piece of timber was joined, to ensure structural stability, the timber was alternately laminated during the lapping process to prevent the joints of the lower timber from overlapping the joints of the upper timber.

4.1.3. Structure Construction

The structural construction of the building is divided into two main parts: the joints in the timber and the connection of the structural units. In the construction of each unit, the timbers of the same layer are not connected, and the timber components are connected between layers by applying glue and nails, which can be performed by robots in the subsequent construction operations. The connection between the units also requires glue and nails, which need to be lifted by a crane and connected manually to complete the construction, as shown in Figure 5. A total of 818 nailing points and 405 gluing points were required for the entire structure.

4.2. Construction Procedure

4.2.1. Determining the Reachable Range of the Robot

The construction of this wooden building required a total of 422 timbers. The length is 5.565 m, the width is 5.775 m, and the height is 5.905 m. Since the building was built by units and each unit was constructed by being placed horizontally, the construction height of each unit is the width of the timber used for construction multiplied by the number of floors. The maximum number of units in the entire construction task is 9 floors and the height of the units is 0.945 m. This requires the manipulable range of the robotic arm to be greater than 5.905 × 5.775 × 0.945 m.

Before starting the fabrication process, the construction range of the robot and the Y–axis slide combination should be determined first. To begin with, a set of points were placed around the robot at 50 mm intervals in the positive direction of the XYZ–axis. Next, the robot reach points (green part) were calculated using KUKA|PRC analysis. Then, all the points on the Z–axis were stacked to 360 degrees, and then all the points on the Z–axis were stacked to 270 degrees. Finally, the cube with the largest side length was extracted from the point grouping, and the reconstructed building cell is located inside this cube, as seen Figure 6.

The combination of the construction robot and the Y-axis chute can effectively complete the construction task of each unit of the target building. The robot places the timber in the construction order from bottom to top according to the construction logic of each unit and applies glue and nail to the necessary timber, as shown in Figure 7.

Each unit is constructed in the same way. After the robot construction of all seven units is completed, each unit is combined to form a complete building, as shown in Figure 8.

4.2.2. Robot Construction Steps

The purpose of using robots in the construction field is to minimize the amount of manual labor in the construction process. Four steps are replaced: picking up timbers, gluing, and placing timbers and nails, as shown in Figure 9.

• Pick

Throughout the robot building process, the robot arm grasps the timber as the first step, and the grasping point set in the robot construction program is the midpoint of the timber. This step requires an operator to provide the timber in real-time and to position the midpoint of the timber at the gripping point of the robotic end effector.

• Glue

The connection between the timber is made using an adhesive connection and a nail joint connection, both of which are performed by a robot. The device for the adhesive connection is a glue box, which consists of a box filled with a waterborne polymer wood binder and a roller on which access to glue is given. The robotic end effector grips the timbers, and its trajectory is along a line segment with the uppermost side of the roller as the midpoint. The length of the glue smear is the longest length of the timbers used in the construction, thus ensuring the effective length of each stick.

• Nail

The robotic end effector also incorporates an air nail gun that fires nails longer than the width of the timber. When the air nail gun fires the nail, the timber is placed in the set position and the gripper remains in the timber’s gripping position. When nailing to the other side, the robot A6 axis rotates 180 degrees and then performs the same operation. The first layer of each unit does not have the glue and nail step.

• Place

In this step, to prevent the timber from colliding with the surrounding timbers in the process of placing, the motion path of the robot arm cannot be the LIN movement (linear movement) method of straight-line motion between two points but needs to be the PTP (point-to-point movement) +LIN movement method through the main rotation of the fifth axis of the robot.

The logic of robotic construction is the same as manual construction in that each unit is built separately in building order. Compared to manual construction, robotic construction requires the digital process of converting each step of the construction instructions into information that can be received by the robot through the software grasshopper (grasshopper version 1. 0. 0005, created by David Rutten at Robert McNeel & Associates in Washington, United States.). In this process, the robot receives the instructions and encounters problems in the process of on-site operation, which can be fed back to the computer to readjust the construction process, to ensure that the robot can complete the construction on site.

The following flowchart outlines this fabrication control loop, which controls the robotic nodes for accomplishing the building’s fabrication (Figure 10).

5. Results and Validation

5.1. Construction Evaluation

Construction evaluation is the focus of this comparative experiment. The advantages and disadvantages of two construction methods of the same timber frame building are compared through both manual and robotic construction. Next, the efficiency of both sides is evaluated mainly in terms of time efficiency evaluation, human efficiency evaluation, and construction quality evaluation.

5.1.1. Time Efficiency Evaluation

Time efficiency is an important reflection of a project’s efficiency, and therefore its evaluation is the focus of this comparison. All other things being equal, increased time efficiency means increased overall efficiency of the project, thus shortening the overall project duration, which in turn saves a range of costs such as labor, site, and equipment. This makes sense for any construction project.

• Manual Construction

Many uncertainties exist in manual construction, which can be affected by uncertainties such as site conditions, number of constructors, and work efficiency. As shown in Figure 11, the shortest construction time in building construction is the sixth layer of unit B, which is about 8 min. In addition, the longest construction time was 53 min for the second layer of Unit E. One of the reasons for this difference is the number of constructors. Unit B has eight people and unit E has three people. This is the main reason for the difference in working hours and shows that the number of people involved in the manual construction process can directly affect the construction time.

Figure 12 shows the total time to build each unit multiplied by the number of constructors in that unit, indicating the sum of the build times of all constructors in each unit. As shown in this figure, unit A is the least efficient and unit E is the most efficient. First, it can be concluded that in this construction, not too many constructors are needed for each unit, and more than a certain number of constructors will affect the overall project construction efficiency. Second, unit A is the first construction unit, and the constructors are not familiar with the construction process, which leads to inefficiency. It can be concluded that not only the number of constructors will affect the construction time, but also the efficiency and proficiency of the constructors will affect the overall construction efficiency.

• Robotic Construction

The robot performs the construction steps in a set sequence within its static working range, which does not require human intervention. The design of the grasshopper takes into account each step of the construction process, so that the robot’s construction time can be calculated directly from the program and is stable, as shown in Figure 13. Only three operators are needed for the entire construction process: two operators place the timbers in the construction order at the pick-up point, and one operator runs the robot.

In the whole construction process, operators need to participate in a limited number of steps, and the construction time is controlled by parametric design. After meeting the number of operators required for construction, continuing to increase the number cannot improve the overall construction time. Thus, it can be seen that the robot construction is stable in construction time and manual use.

The comparison shows that the robot construction time is shorter than the manual construction time, is more efficient and stable, and the efficiency of the robot in its normal construction state is not affected by the number of construction workers, as shown in Figure 14. Therefore, robot construction has an important impact on improving the efficiency of construction time.

5.1.2. Construction Quality Evaluation

Construction quality evaluation is equally important for construction projects. In addition to high construction efficiency, construction accuracy is also the key to measuring the merits of a project.

• Manual Construction

Due to the level of construction workers, the current situation of the building site, the precision of construction tools, and other aspects of the impact of on-site construction, there will be a certain degree of construction errors, and the accumulation of errors layer by layer will cause some interference with the overall construction quality, reducing the accuracy of construction. In the case of manually constructed buildings, errors may exist in every step from timber cutting to timber building and timber joining.

The chainsaw gears used to cut the timber are very thick and manually operated, so the cutting results will deviate from the required dimensions. As shown in Figure 15a, in the measurement of the timber after cutting, the actual cut size of the timber with the design size of 2900 mm is 2896 mm and the error has started to appear.

There are also deviations in the process of erection and connection of timber, and the connection gaps tend to appear in the articulated part of the building facade and roof, as shown in Figure 15b, where the facade is displaced during the erection process, making the wall surface not flat enough. All the above reasons caused the reduction in building construction accuracy.

• Robot Construction

Design methods and design tools have become increasingly complex, variable, and data-dependent in the pursuit of architectural design accuracy, and this accuracy is often not fully implemented in the process of translation and construction because the construction tools do not dovetail with this data-based mode of operation. The robotics platform makes up for this aspect of lack, making the design and high-precision construction of buildings truly a set of the continuous and complete model. In contrast to simulation experiments, the actual construction process of the robot will still have errors beyond the design, which requires stopping the construction experiment and adjusting the errors in the program, modifying the construction program, and then continuing the previously interrupted construction. This is considered to be the robot’s real-time adjustment and correction of construction errors, which allows it to complete the construction with higher accuracy.

The picture shown in Figure 16 is the robot construction site. It can be seen that the robot construction process has less error and higher accuracy, and the constructed building better meets the design and construction requirements, which means that the whole construction process is of higher quality.

5.2. Validation

The robot arm length in the laboratory is 900 mm, which is limited by its model and cannot be used for the 1:1 scale building construction in the previous simulation experiment. Therefore, the practical building experiment was built by the robot at a 1:5 scale of the original building, as shown in Figure 16. The practical construction used the same parametric construction tasks, and the robot received and performed the same construction tasks, aiming to verify the rationality and implement ability of the construction logic and procedures of the previous simulation experiment. In addition, the rationality of the efficiency of the simulation experiment was verified based on the practical construction experiment time.

As shown in

Table 1. The proportional construction time difference between the simulation experiment and the practical construction.

Unit	A	B	C	D	E	F	G
1:1 Scale simulation experiment time (Min)	89.8	105.9	94.8	87.8	67.0	77.6	64.7
1:5 Scale practical construction time (Min)	34.9	40.4	37.4	34.0	25.7	29.8	24.7
The proportion of construction time difference	2.57	2.62	2.54	2.58	2.61	2.60	2.62

, the construction time for each unit of the building is shown for two construction scales, and due to the different construction scales, the construction time for the 1:5 scale practical construction is less than the construction time for the 1:1 scale simulation experiment under the same construction procedure. However, the proportional construction time difference between the simulation experiment and the practical construction is always within the range of 2.54–2.62, which is very stable, indicating that the construction time is relatively stable under the same parametric construction task, and it can be verified that the construction time measured by the equal proportional simulation experiment is similar to the real-time practical construction.

6. Discussion

A multidisciplinary approach combining robotics and building construction has made it possible to conduct experiments using arm robots for the construction of target buildings. The successful implementation of the experiment marks the potential for integrating parametric design and robotic construction into a wood building construction project [26]. In addition, by comparing manual construction with the experimental process and experimental results, it is verified that robotic construction contributes to the improvement of construction efficiency and the reduction in labor use and material waste, thus improving the environmental and economic benefits of the whole construction process and achieving sustainability of construction. The accomplishments of the experiment can be summarized as follows: First, multiple construction steps of the building can be accomplished by parametric design operating the end effector of the robot. Repeated construction steps of individual timbers are completely built by the robot. Second, it is verified that robotic construction highlights its high efficiency, stability, and accuracy compared to manual construction. Third, it is demonstrated that the logical workflow of parametric design can easily combine interdisciplinary design and construction techniques for testing, generation, and analysis in a digital environment. The completion of simulation and actual experiments also signifies that the possibility of bringing robots into whole-house construction is available. Compared to the literature and cases summarized in the previous sections, this paper explores the design and construction process of whole-house wooden construction by an arm robot based on robotic structural construction, robotic architectural component manufacturing, and robotic 3D printing of buildings, and hopes to provide a basis for future marketable robotic construction.

For the future of robotic wood construction, it is also clear from the experiments that: first, the size of the stationary robot determines the scale of the building that can be constructed, and the feasibility of mobile robot construction needs to be considered for the construction of larger buildings in the future; second, the connection mode of wood is not only glue and nail but also different connection modes for different structural types of wood buildings, so the adaptability between different connection modes and robot operation needs to be considered; and third, the robot’s end effector needs to match the application steps of the building construction, which means that the development of the end effector is equally important for the future development of the robot construction.

7. Conclusions

The study of robotic timber building construction, especially for the whole building, is not yet complete and is still in its infancy. The application of robotics poses many theoretical and practical challenges to traditional architectural science based on socially justifiable research. For scientific research, this paper combined experimental and theoretical methods in the hope of verifying that the robotic construction process, which is an integrated process of digital design and automated manufacturing, allows traditional wood construction to fundamentally expand its scope, change the construction process, and introduce the assembly logic of robotic construction into the wood construction industry [11]. This paper even proposed and attempted with experimentation a digital construction strategy for whole-house construction, yet the future of robotic construction of wood buildings is still full of challenges. However, the research was meaningful: robotic construction not only creates a new vision for the future of wood construction but also highlights new possibilities to explore and understand it [16].