- Email: [email protected]
Development of virtual and real-field construction management systems in innovative, intelligent field factory
Automation in Construction 9 Ž2000. 503–514 www.elsevier.comrlocaterautcon
Development of virtual and real-field construction management systems in innovative, intelligent field factory Yoshio Maruyama a,) , Yoichiro Iwase a , Kazuo Koga b, Junichi Yagi c , Hiroo Takada c , Naohisa Sunaga c , Shigeomi Nishigaki d , Takashi Ito e, Kinya Tamaki f a
f
Technical Research Institute, Hazama Corp., 515-1 Nishi-mukai Karima, Tsukuba, Ibaraki 305-0822, Japan b Building Construction DiÕision, Hazama Corp., 2-5-8 Kita-Aoyama, Minato-Ku, Tokyo 107-8658, Japan c Institute of Technology, Shimizu Corp., 4-17, Etchujima 3-choume, Koto-Ku, Tokyo 135-8530, Japan d Kick Ltd., 5-34-7 Honcho, Nakano-Ku, Tokyo 164-0012, Japan e Niigata System Technologies Inc., 1-9-3, Kamata-Honcho, Ohta-Ku, Tokyo 144-8640, Japan School of Business Administration, Aoyama Gakuin UniÕersity, 4-4-25 Shibuya-Ku, Tokyo 150-8836, Japan
Abstract In this study, we proposed a concept of virtual and real-field construction management systems ŽVR-Coms., which is integrated with virtual construction simulation, planning, scheduling, and performance management systems to evaluate productivity and safety in virtual simulated and real-field constructions. And, we built up a computational environment to develop the VR-Coms. The VR-Coms offer supporting modules for learning and discovering solutions with objective to manage construction at right speed with improved humanware and constructability. The configuration of VR-Coms is described. This paper also shows an application of agent theory to construction management. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Construction management; Planning; Scheduling; Construction simulation; Collaboration
1. Introduction We developed virtual and real-field construction management systems ŽVR-Coms., which are integrated with virtual construction simulation, planning, scheduling, and performance management systems to evaluate productivity and safety in virtual simulated )
Corresponding author. E-mail addresses: [email protected] ŽY. Maruyama., [email protected] ŽY. Iwase., [email protected] ŽK. Koga., [email protected] ŽJ. Yagi., [email protected] ŽH. Takada., [email protected] ŽN. Sunaga., [email protected] ŽS. Nishigaki., [email protected] ŽT. Ito., [email protected] ŽK. Tamaki..
and real-field constructions. The VR-Coms offer supporting modules for learning and discovering solutions based on previous experience and performance evaluation with the objective to manage construction at right speed with improved humanware and constructability. The function of leadership, followership, and reciprocal interaction between leaders and followers are collectively termed as humanware w1–3x. Constructability is defined as the optimum use of construction knowledge and experience in planning, engineering, procurement, and field operations to achieve overall project objective w4x. ‘‘Innovative, Intelligent Field Factory ŽIF7.’’ is one project of Intelligent Manufacturing System
0926-5805r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 5 8 0 5 Ž 0 0 . 0 0 0 6 1 - 3
504
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
ŽIMS. program in Japan. IMS program was proposed by Japan in 1989. The objectives of IMS program are to provide an exciting opportunity for international cooperation in advanced manufacturing, and a vision and structure for worldwide sharing of manufacturing technology development with cost, risks, and benefits in a balanced and equitable manner. The IF7 project consists of three work packages as follows: Ž1. Work package 1: Assembly methods for large-scale structures. Particular emphasis will be placed on realization of modularized construction system by utilizing flexible advanced manufacturing technologies. Ž2. Work package 2: Mechanization systems for assemble of large-scale structures. Automated systems are studied such as technology of heavy material transportation, 3D position detection system, and a theory of construction material handling method. Ž3. Work package 3: VR-Coms. VRComs in innovative, intelligent field factory is studied and developed. This package has four tasks as follows: Task 1: Study on real-time based concurrent construction management systems to add value by enhancing safety and productivity in overall works from construction planning to field factory operations, Task 2: Development of construction simulation systems using VR technique integrated with planning, scheduling, and performance management systems, Task 3: Study on field factory transparency, where detailed information, proactive advises and guidelines in all locations are accessible electronically by workers for thinking and decision making, and Task 4: Study on application of agent theory to safety management, resources management, facility management, and large-scale distributed information management. This paper reports on the development of VRComs as part of the research of IF7.
2. Concept of VR-Coms Fig. 1 shows the schematic view of VR-Coms. The VR-Coms has two construction fields, one is
virtual construction field, and the other is real construction field. The upper row of Fig. 1 shows the concept of virtual construction field. In this field, information technology ŽIT. is made good use of and information on the building design and the scheme of execution is generated before real construction is started. On the other hand, the lower shows the real construction field. Here, parties concerned necessary for constructing real buildings are shown, such as construction management office, site factory, subcontractor, materials suppliers, equipment suppliers, machine parts suppliers, and architect offices. VR-Coms can be used for site manager to study construction planning and management. Moreover, it is possible to use VR-Coms to support construction planning and to manage construction projects. Experienced and skilled workers review and modify all construction plans and schedules before the implementation of plans. Typically, an iterative procedure is taken to develop solution to accommodate resource constraints, including the workers and equipment involved in the project. In this iterative process, alternatives are generated and tested for feasibility. Further improvements or detection of potential problems are studied to generate new alternatives. This iterative generate-and-test strategy can be applied to evaluate alternative plans and schedules until a satisfactory plan with an appropriate schedule is obtained. The purpose of virtual construction field integrated with planning and scheduling system in VR-Coms is to facilitate the execution of generateand-test strategy.
3. Development of test bed of VR-Coms To develop VR-Coms, a computerized environment as shown in Fig. 2 was built up within the School of Business Administration, Aoyama Gakuin University. The computerized environment is composed of the following functions: 1. 2. 3. 4. 5.
Design function, Planning and Scheduling function, Line simulation function, Work simulation function, Real-time based construction management and performance management function,
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
505
Fig. 1. Schematic view of VR-Coms.
6. Database and Library, and 7. Cyber agents.
function cooperate mutually. Cyber agents help site engineers to use VR-Coms.
The design function and the planning and scheduling function generate construction plans and schedules. The line simulation and the work simulation function test the construction plans and schedules. Real-time based management function monitors the output data of simulation, and converts to some information profitable for construction management. The site engineers update the first plan and schedule referring to the result of test. The database makes the
3.1. Design function The role of the design function is to make two dimension drawings, and to make building model necessary for construction planning and execution of virtual construction simulation. The building model contains not only three dimension shapes but also all related data such as materials of structure, amount of parts, cost, etc.
506
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
5. Part composition element 2: Parts such as sanitary wares and chairs. 6. Space composition element: This is composed of combination with the basic composition element. This element is classified into external space, buffer space, and interior space.
3.2. Planning and scheduling function The objectives of planning and scheduling function are to make a construction schedule and plan based on the master building plan, template of network diagram, time required each work and the restriction condition of building construction project. This function makes the process plan over the entire construction project. Moreover, the daily work schedule of every day and the work plan of the construction machines should be able to be made by this function. The earliest and latest node time of each node of construction schedule can be calculated by substitute the time required of each work, the beginning date of the project, and the completion date of the project in the network diagram. In addition, four kinds of time to each work can be calculated by the following equations: Fig. 2. Computerized environment to develop VR-Coms.
Earliest starting time s earliest node time,
Ž 1.
Earliest finishing time s earliest node time q duration of work, A basic composition element of the building is the following six elements w5x. The design function should decide details of the following elements concerning an individual building. 1. Base element: Semantics of physical environment like plan site, road, adjoining land, etc. 2. Basic composition element: Main semantic group, which composes spaces such as wall, pillars, and roofs. Basic composition elements form main structure of building. 3. Opening composition element: This elements are fittings in door and window, etc., and simple openings. 4. Part composition element 1: Machines such as elevators, escalators, and stairs.
Ž 2.
Latest starting time s latest node time y duration of work, Latest finishing time s latest node time.
Ž 3. Ž 4.
A critical path of the work schedule is clarified with use of these four kinds of time and the network diagram. If the number of worker and a resource necessary for each work are input, necessary number of worker for the construction project and the amount of the resource can be estimated. The output obtained from the planning and scheduling function are the following: 1. Work schedule which resource allocation is completed,
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
2. Earliest starting time, earliest finishing time, latest starting time, latest finishing time, and float of each work, 3. Critical path of the project, 4. Total cost of the project, and 5. Construction site layout including location temporary housing, construction machine, and transporting rout of materials. 3.3. Line simulation function The line simulation function numerically simulates material flow in the construction site. The purpose of line simulation is to confirm whether the construction schedule generated by planning and scheduling function is actually executable. And, the line simulation forecasts whether an undesirable phase appears or not by calculating a transportation load in the schedule. The following items can be examined by the line simulation function: 1. Through put, 2. Capacity of temporary depository and its operation, 3. Order of turning on materials and turning on time, 4. Layout of construction site including site factory, 5. Estimation of influence by disturbance and examination of countermeasures, The construction schedule and the temporaryhousing plan are necessary to execute the line simulation. The line simulation function generates the following results: 1. 2. 3. 4.
Validity checked construction site layout, Presumption value of through put, Storage capacity and its operation method, The order of turning on materials and turning on time, 5. Countermeasures against turbulence. 3.4. Work simulation function The work simulation function in VR-Coms creates virtual construction in 3D computer-simulated environment before the planned construction is put
507
into practice. The virtual construction simulation system is integrated with planning, scheduling, and performance management function, so that we will be able to find potential hazards latent in the plan being considered and to evaluate productivity. The 3D models of building, construction site, other machines, and tools to be used in the construction activity are necessary to execute the work simulation. The 3D building model comes from the design function, and model of construction site is generated by planning function. 3D models of construction machines and tools are saved in library. Moreover, work description, work plan, and operation plan of workers and machines are needed to run the work simulation. The engineer should make these data before executing the simulation. The following outputs are obtained by simulation: 1. Time necessary for work, 2. Place with which some one spatially interferes, 3. Verified control algorithm of construction equipment, 4. Verified transportation path, and 5. Verified work traffic line. Fig. 3 shows the flow of the planning and scheduling process, which uses each function of work simulation, line simulation, and planning and scheduling. 3.5. Real-time based construction management and performance management function The real-time based construction management and performance management system monitors the behavior of the virtual construction simulation and captures the data that need to analyze productivity and safety of construction activity. The purpose of real-time based construction management system is to reduce manufacturing overheads and bears the third profit. The real-time based construction management system should correctly present the situation of construction site to site managers. The system converts monitored data into the following information and offers it to the site manager: 1. Information for production control, 2. Information for cost control,
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
508
refer to the state of the virtual and real-field in the past, present, or future, and to the mental state of other computerized agents. Capabilities are taken to be a primitive notion defined in terms of futurebranching structures. Commitment rules are activated based on certain patterns in an incoming message and current mental condition, that is, the mental state of this computerized agent. When the time comes to execute the action, the mental state at the time will be examined to see whether the mental condition is satisfied. Generally, computerized agents may be characterized in terms of autonomous, intelligent, communication, and anthropomorphic. In this research project, the following cyber-agents are considered: 1. Application agent, Ža. Control-recovery agents, Žb. Fault-warning agents, Žc. Procedure-knowledge agents, and Žd. Data mining agents, and 2. Network agents.
4. Current research activities Fig. 3. Construction planning flow using VR-Coms.
3. Information for quality control, 4. Information for operation of equipment machine, and 5. Information for maintenance of equipment machine.
3.6. Database Main contents of database are as-designed data, as-built data, and library that can be used together by all projects. The data generated by a function is passed to another function via the database.
Current research activities are the following: 1. Control-recovery agent and fault-warning agent for automated construction operations; 2. Virtual dynamic construction simulation: Ža. Case study on Takada model of new construction method for multi-stories apartment building, Žb. Construction line simulation, and Žc. Computational models to evaluate safety and productivity; 3. Improvement of the computerized test bed environment toward the VR-Coms; and 4. Object-oriented database model with respect to building construction.
3.7. Cyber agents 4.1. Case study of control-recoÕery agent The term ‘‘agent’’ means someone acting on behalf of someone else. A computerized agent is an entity who has own mental state being composed of beliefs, capabilities, and commitment w7x. Beliefs
We executed a case study of agentification of control-recovery aids in the control, monitoring and warning computerized system for a semi-automated
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
sliding system of assembled roof within a underground power plant w8x. Fig. 4 shows schematic view of the prototype of control-recovery agents being developed in the case. The control, monitoring and warning computerized system here is functioning to: 1. Monitor the behavior of the assembled roof being slid, 2. Check whether its behavior is approaching its safety operating limits under operations, and 3. Automatically stop its behavior in emergency. The control-recovery agents translate the raw sensory data associated with the key variables into control or error signals, signs for displaying key warning and taking proactive actions, and symbols w6x for reasoning the underlying causes in order to
509
invoke error recovery functions at skill-, rule- and knowledge-based levels w6x. Signal is continuous quantitative indicator of the time-space behavior of the semi-automatic sliding system. The signal is categorized into control signal and error signal w6x. The control signal shows the transient state of behavior at a specific point in time. The error signal represents the differences between the actual state and the intended or planned state in a time-space behavior. The control-recovery agent stores a notion of the boundaries of acceptable performance Že.g., safety control limits. and the obligation to activate emergency shut down function if values of signal go beyond safety control limits in the capability. Workers can understand the existing state of the system by the control signal and off-normal state by the error signal. Signs are labeled by
Fig. 4. Schematic view of prototype of control-recovery agents.
510
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
names of states or situations of the semi-automatic sliding system w6x. Signs, which are built in IF–THEN rules, are used to select or activate stored predetermined actions that control the sequence of routine operations. Cyber-agent can display the key warnings in emergency and evoke built-in self-stopping and off-normal inspection functions when the raw sensory data goes beyond the control limit value. Workers can select and activate predetermined recovery actions by signs. Symbols refer to concepts tied to functional properties and can be used for reasoning and computation by means of a suitable representation of such property w6x. Since workers’ attention are likely to be directed at a small part of the total problem space at any one time, mimic diagram is relevant for them to direct their attention to logically important aspects of the problem space.
Information is treated as symbols in the mimic diagram to topographically investigate improper function and to symptomatically reason its underlying causes. The control-recovery agent here can denote the improper function topographically by making its location brightness and blinking in the mimic diagram. Workers can reason some potential faults based on symbols, as confirming and asking for additional data. 4.2. Case study on coffee-table type of construction method 4.2.1. Outline of Takada model Shimizu in Japan proposes a new construction method called coffee-table construction method for multi-stories apartment building. Fig. 5 shows
Fig. 5. Virtual dynamic construction with coffee-table method.
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
overview of building being constructed by this construction method. In this construction method, coffee-table modules are assembled in site factory and are transported and erected in order at the installation area. This method increases productivity and shortens construction duration compared with current construction method under stable conditions. Because we can assemble coffee-table modules under safety and invariable environment in site factory. In this case study, created is a virtual dynamic construction of the Takada model in the computerized test bed environment. 4.2.2. Outline of site factory As for division of main structure such as column, beam and load-bearing wall, details of the division
511
shall be standardized as considering conditions of design and construction method as a whole. The coffee-table construction method is composed of subsystems as for: 1. 2. 3. 4.
Transportation from supplier to site factory, Assembly works at site factory, Hoisting operation, and Operations of installation and connection. Job coordination is conducted by work rules among the subsystems and of themselves.
In this construction method, coffee-table modules are assembled at site factory and transported to a final installation position. At present, reasonable hoisting machine is tower crane. Size and weight of
Fig. 6. Virtual line simulation with coffee-table method.
512
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
the coffee-table module shall be designed as considering capacity of tower crane. Parts on truck are delivered from supplier to site factory. The parts are unloaded and stockpiled by hoist machine at unloading area. Hoist traveler grasps and transport the each part and select a rail to designated position according to the ID of part. As for subsystem at site factory, the following items shall be studied to develop optimal facility. Ø Number of rails for unloading and assembling, Ø Number of hoist traveler, Ø Precedence relationships between operations of machinery and equipment, Ø Conditions of entrance for trucks, Ø Assembly time for parts,
Ø Erection time for units, and Ø Bottleneck in process.
4.2.3. Production line simulation for coffee-table type of construction method This virtual dynamic construction centers on: 1. Multilevel assembly sequence ( Tooling operation ( Assembly operation ( Delay operation 2. Interference checks and Collision avoidance ( Part trajectory paths ( Swept volume envelope ( Dynamic cross-sectioning.
Fig. 7. Improvement of the computerized test bed environment.
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
513
Simulating the Takada model in the computerized test bed environment enables us to do research on:
calculated. The length of the time depends on each following item:
1. Planning: Determine work that can be released during certain time interval; 2. Scheduling: Determine work when that work is released into the field factory on a-minute-byminute basis; 3. Status and clock time: Know work position and machine status as well as ensuring that the plan representation covers the current clock time up to some predefined future plan horizon; and 4. What-if: ( Accept or reject the resulting plan; and ( Explore the consequence of changes to the current state of the field factory.
Ø The order of turning on PCa parts, Ø The time when PCa parts are turned on, Ø Layout of site factory and tower crane in construction site, Ø Transportation routes of PCa parts, Ø Ability of tower crane, and Ø Productivity of site factory.
Fig. 6 shows an example of line simulation of construction site where the coffee-table type of construction method is adapted. With this simulation, time necessary to build one layer of the building was
The combination of these items which satisfies a basic building plan can be obtained by repeating the line simulation with various conditions. We can improve the first construction plan and schedule referring to the result of the line simulation. 4.3. ImproÕement of the computerized test bed enÕironment The virtual construction simulation system will be able to create virtual construction in 3D computer-
Fig. 8. Configuration of VR-Coms.
514
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
simulated environment before the planned construction is put into practice. The virtual construction simulation system is integrated with a planning, scheduling and performance management system, so that we will be able to find potential hazards latent in the plan being considered and to evaluate productivity. Fig. 7 shows mechanism to be built in the computerized test bed environment. Each task involved in construction works is grouped into the work packages, so that their forms can be identified in the 3D simulation and their progress can be monitored. The planning and scheduling systems are employed to produce the information regarding the work packages on the project and to determine when each work package is dispatched to the construction site. With the planning module and virtual simulation tools, ‘‘what-if’’ analyses can be carried out to explore what happens in the construction site, as well as when and where the events occur. The virtual simulation, when combining with the scheduling system, can be used to study the status of the project and to relate the work performed at a specific time and specific location. The control, monitoring and warning system monitors the behavior of the virtual construction simulation and captures the data needed to analyze productivity and safety. The coffee-table construction method being simulated in and this improvement of the computerized test bed environment will configure the VR-Coms as shown in Fig. 8.
5. Conclusion In this paper, the concept of VR-Coms was proposed and computerized test bed was built up. The computerized test bed of VR-Coms is integrated with virtual construction simulation, planning and scheduling, and real-time based construction management functions. First, we made the prototype of Control-recovery Agent on the test bed and attempted the operation confirmation of it. Second, we tried making the production plan of building work which used the coffee-table construction method. As a result, it has been understood that the test bed is useful as a
development environment of cyber-agent. We confirmed that the virtual construction simulation system integrated with planning, scheduling, and performance management function facilitate achieving interactive generate-and-test strategy to improve construction plans and schedules. Future work will concentrate on the configuration of VR-Coms improving the test bed environment. The anticipated results that this study incubates are summarized as follows: 1. Purchaser’s merits ( The content of the order is enhanced ( The design change is flexible ( Term of works shortening ( Reduction of construction expense 2. Cooperation trader’s merits ( An increase in unexpected construction expense is prevented ( The construction delay is prevented ( The arrangement to expect the big factor of safety becomes useless 3. Designer and contractor’s merits ( The design change is flexible ( Term of works shortening ( Reduction of construction expense.
References w1x S. Nishigaki, The application of discriminate analysis to safety patrol system, report of statistical application research, Union Jpn. Sci. Eng. 35 Ž1. Ž1988. 24–34. w2x S. Nishigaki et al., Humanware, human error, and Hiyari-hat: a template of unsafe symptoms, J. 1 of construction engineering and management, ASCE 120 Ž2. Ž1994. 421–442. w3x S. Nishigaki et al., Safety problems in on-site construction work processes, in: Proc. 11th ISARC, Brighten, U.K., 1997, pp. 13–18. w4x Construction Industry Institute Publication Nos: 3-1-Constructability: A Primer, July 1986. w5x Creation of building and urban city by intelligent system, Gihodo, 1998 ŽIn Japanese.. w6x J. Rasmussen, Skills, rules, and knowledge; signals, signs, and symbols, and other distinctions in human performance models, IEEE Trans. SMC SMC-13 Ž3. Ž1983. 257–266. w7x Y. Shoham, Agent-oriented programming, Artif. Intell. 60 Ž1993. 51–92. w8x S. Nishigaki et al., Control, monitoring and warning computerized system for a semi-automated sliding system of assembled roof, in: Proc. 14th ISARC,1997, pp. 199–205.
Development of virtual and real-field construction management systems in innovative, intelligent field factory Yoshio Maruyama a,) , Yoichiro Iwase a , Kazuo Koga b, Junichi Yagi c , Hiroo Takada c , Naohisa Sunaga c , Shigeomi Nishigaki d , Takashi Ito e, Kinya Tamaki f a
f
Technical Research Institute, Hazama Corp., 515-1 Nishi-mukai Karima, Tsukuba, Ibaraki 305-0822, Japan b Building Construction DiÕision, Hazama Corp., 2-5-8 Kita-Aoyama, Minato-Ku, Tokyo 107-8658, Japan c Institute of Technology, Shimizu Corp., 4-17, Etchujima 3-choume, Koto-Ku, Tokyo 135-8530, Japan d Kick Ltd., 5-34-7 Honcho, Nakano-Ku, Tokyo 164-0012, Japan e Niigata System Technologies Inc., 1-9-3, Kamata-Honcho, Ohta-Ku, Tokyo 144-8640, Japan School of Business Administration, Aoyama Gakuin UniÕersity, 4-4-25 Shibuya-Ku, Tokyo 150-8836, Japan
Abstract In this study, we proposed a concept of virtual and real-field construction management systems ŽVR-Coms., which is integrated with virtual construction simulation, planning, scheduling, and performance management systems to evaluate productivity and safety in virtual simulated and real-field constructions. And, we built up a computational environment to develop the VR-Coms. The VR-Coms offer supporting modules for learning and discovering solutions with objective to manage construction at right speed with improved humanware and constructability. The configuration of VR-Coms is described. This paper also shows an application of agent theory to construction management. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Construction management; Planning; Scheduling; Construction simulation; Collaboration
1. Introduction We developed virtual and real-field construction management systems ŽVR-Coms., which are integrated with virtual construction simulation, planning, scheduling, and performance management systems to evaluate productivity and safety in virtual simulated )
Corresponding author. E-mail addresses: [email protected] ŽY. Maruyama., [email protected] ŽY. Iwase., [email protected] ŽK. Koga., [email protected] ŽJ. Yagi., [email protected] ŽH. Takada., [email protected] ŽN. Sunaga., [email protected] ŽS. Nishigaki., [email protected] ŽT. Ito., [email protected] ŽK. Tamaki..
and real-field constructions. The VR-Coms offer supporting modules for learning and discovering solutions based on previous experience and performance evaluation with the objective to manage construction at right speed with improved humanware and constructability. The function of leadership, followership, and reciprocal interaction between leaders and followers are collectively termed as humanware w1–3x. Constructability is defined as the optimum use of construction knowledge and experience in planning, engineering, procurement, and field operations to achieve overall project objective w4x. ‘‘Innovative, Intelligent Field Factory ŽIF7.’’ is one project of Intelligent Manufacturing System
0926-5805r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 5 8 0 5 Ž 0 0 . 0 0 0 6 1 - 3
504
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
ŽIMS. program in Japan. IMS program was proposed by Japan in 1989. The objectives of IMS program are to provide an exciting opportunity for international cooperation in advanced manufacturing, and a vision and structure for worldwide sharing of manufacturing technology development with cost, risks, and benefits in a balanced and equitable manner. The IF7 project consists of three work packages as follows: Ž1. Work package 1: Assembly methods for large-scale structures. Particular emphasis will be placed on realization of modularized construction system by utilizing flexible advanced manufacturing technologies. Ž2. Work package 2: Mechanization systems for assemble of large-scale structures. Automated systems are studied such as technology of heavy material transportation, 3D position detection system, and a theory of construction material handling method. Ž3. Work package 3: VR-Coms. VRComs in innovative, intelligent field factory is studied and developed. This package has four tasks as follows: Task 1: Study on real-time based concurrent construction management systems to add value by enhancing safety and productivity in overall works from construction planning to field factory operations, Task 2: Development of construction simulation systems using VR technique integrated with planning, scheduling, and performance management systems, Task 3: Study on field factory transparency, where detailed information, proactive advises and guidelines in all locations are accessible electronically by workers for thinking and decision making, and Task 4: Study on application of agent theory to safety management, resources management, facility management, and large-scale distributed information management. This paper reports on the development of VRComs as part of the research of IF7.
2. Concept of VR-Coms Fig. 1 shows the schematic view of VR-Coms. The VR-Coms has two construction fields, one is
virtual construction field, and the other is real construction field. The upper row of Fig. 1 shows the concept of virtual construction field. In this field, information technology ŽIT. is made good use of and information on the building design and the scheme of execution is generated before real construction is started. On the other hand, the lower shows the real construction field. Here, parties concerned necessary for constructing real buildings are shown, such as construction management office, site factory, subcontractor, materials suppliers, equipment suppliers, machine parts suppliers, and architect offices. VR-Coms can be used for site manager to study construction planning and management. Moreover, it is possible to use VR-Coms to support construction planning and to manage construction projects. Experienced and skilled workers review and modify all construction plans and schedules before the implementation of plans. Typically, an iterative procedure is taken to develop solution to accommodate resource constraints, including the workers and equipment involved in the project. In this iterative process, alternatives are generated and tested for feasibility. Further improvements or detection of potential problems are studied to generate new alternatives. This iterative generate-and-test strategy can be applied to evaluate alternative plans and schedules until a satisfactory plan with an appropriate schedule is obtained. The purpose of virtual construction field integrated with planning and scheduling system in VR-Coms is to facilitate the execution of generateand-test strategy.
3. Development of test bed of VR-Coms To develop VR-Coms, a computerized environment as shown in Fig. 2 was built up within the School of Business Administration, Aoyama Gakuin University. The computerized environment is composed of the following functions: 1. 2. 3. 4. 5.
Design function, Planning and Scheduling function, Line simulation function, Work simulation function, Real-time based construction management and performance management function,
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
505
Fig. 1. Schematic view of VR-Coms.
6. Database and Library, and 7. Cyber agents.
function cooperate mutually. Cyber agents help site engineers to use VR-Coms.
The design function and the planning and scheduling function generate construction plans and schedules. The line simulation and the work simulation function test the construction plans and schedules. Real-time based management function monitors the output data of simulation, and converts to some information profitable for construction management. The site engineers update the first plan and schedule referring to the result of test. The database makes the
3.1. Design function The role of the design function is to make two dimension drawings, and to make building model necessary for construction planning and execution of virtual construction simulation. The building model contains not only three dimension shapes but also all related data such as materials of structure, amount of parts, cost, etc.
506
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
5. Part composition element 2: Parts such as sanitary wares and chairs. 6. Space composition element: This is composed of combination with the basic composition element. This element is classified into external space, buffer space, and interior space.
3.2. Planning and scheduling function The objectives of planning and scheduling function are to make a construction schedule and plan based on the master building plan, template of network diagram, time required each work and the restriction condition of building construction project. This function makes the process plan over the entire construction project. Moreover, the daily work schedule of every day and the work plan of the construction machines should be able to be made by this function. The earliest and latest node time of each node of construction schedule can be calculated by substitute the time required of each work, the beginning date of the project, and the completion date of the project in the network diagram. In addition, four kinds of time to each work can be calculated by the following equations: Fig. 2. Computerized environment to develop VR-Coms.
Earliest starting time s earliest node time,
Ž 1.
Earliest finishing time s earliest node time q duration of work, A basic composition element of the building is the following six elements w5x. The design function should decide details of the following elements concerning an individual building. 1. Base element: Semantics of physical environment like plan site, road, adjoining land, etc. 2. Basic composition element: Main semantic group, which composes spaces such as wall, pillars, and roofs. Basic composition elements form main structure of building. 3. Opening composition element: This elements are fittings in door and window, etc., and simple openings. 4. Part composition element 1: Machines such as elevators, escalators, and stairs.
Ž 2.
Latest starting time s latest node time y duration of work, Latest finishing time s latest node time.
Ž 3. Ž 4.
A critical path of the work schedule is clarified with use of these four kinds of time and the network diagram. If the number of worker and a resource necessary for each work are input, necessary number of worker for the construction project and the amount of the resource can be estimated. The output obtained from the planning and scheduling function are the following: 1. Work schedule which resource allocation is completed,
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
2. Earliest starting time, earliest finishing time, latest starting time, latest finishing time, and float of each work, 3. Critical path of the project, 4. Total cost of the project, and 5. Construction site layout including location temporary housing, construction machine, and transporting rout of materials. 3.3. Line simulation function The line simulation function numerically simulates material flow in the construction site. The purpose of line simulation is to confirm whether the construction schedule generated by planning and scheduling function is actually executable. And, the line simulation forecasts whether an undesirable phase appears or not by calculating a transportation load in the schedule. The following items can be examined by the line simulation function: 1. Through put, 2. Capacity of temporary depository and its operation, 3. Order of turning on materials and turning on time, 4. Layout of construction site including site factory, 5. Estimation of influence by disturbance and examination of countermeasures, The construction schedule and the temporaryhousing plan are necessary to execute the line simulation. The line simulation function generates the following results: 1. 2. 3. 4.
Validity checked construction site layout, Presumption value of through put, Storage capacity and its operation method, The order of turning on materials and turning on time, 5. Countermeasures against turbulence. 3.4. Work simulation function The work simulation function in VR-Coms creates virtual construction in 3D computer-simulated environment before the planned construction is put
507
into practice. The virtual construction simulation system is integrated with planning, scheduling, and performance management function, so that we will be able to find potential hazards latent in the plan being considered and to evaluate productivity. The 3D models of building, construction site, other machines, and tools to be used in the construction activity are necessary to execute the work simulation. The 3D building model comes from the design function, and model of construction site is generated by planning function. 3D models of construction machines and tools are saved in library. Moreover, work description, work plan, and operation plan of workers and machines are needed to run the work simulation. The engineer should make these data before executing the simulation. The following outputs are obtained by simulation: 1. Time necessary for work, 2. Place with which some one spatially interferes, 3. Verified control algorithm of construction equipment, 4. Verified transportation path, and 5. Verified work traffic line. Fig. 3 shows the flow of the planning and scheduling process, which uses each function of work simulation, line simulation, and planning and scheduling. 3.5. Real-time based construction management and performance management function The real-time based construction management and performance management system monitors the behavior of the virtual construction simulation and captures the data that need to analyze productivity and safety of construction activity. The purpose of real-time based construction management system is to reduce manufacturing overheads and bears the third profit. The real-time based construction management system should correctly present the situation of construction site to site managers. The system converts monitored data into the following information and offers it to the site manager: 1. Information for production control, 2. Information for cost control,
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
508
refer to the state of the virtual and real-field in the past, present, or future, and to the mental state of other computerized agents. Capabilities are taken to be a primitive notion defined in terms of futurebranching structures. Commitment rules are activated based on certain patterns in an incoming message and current mental condition, that is, the mental state of this computerized agent. When the time comes to execute the action, the mental state at the time will be examined to see whether the mental condition is satisfied. Generally, computerized agents may be characterized in terms of autonomous, intelligent, communication, and anthropomorphic. In this research project, the following cyber-agents are considered: 1. Application agent, Ža. Control-recovery agents, Žb. Fault-warning agents, Žc. Procedure-knowledge agents, and Žd. Data mining agents, and 2. Network agents.
4. Current research activities Fig. 3. Construction planning flow using VR-Coms.
3. Information for quality control, 4. Information for operation of equipment machine, and 5. Information for maintenance of equipment machine.
3.6. Database Main contents of database are as-designed data, as-built data, and library that can be used together by all projects. The data generated by a function is passed to another function via the database.
Current research activities are the following: 1. Control-recovery agent and fault-warning agent for automated construction operations; 2. Virtual dynamic construction simulation: Ža. Case study on Takada model of new construction method for multi-stories apartment building, Žb. Construction line simulation, and Žc. Computational models to evaluate safety and productivity; 3. Improvement of the computerized test bed environment toward the VR-Coms; and 4. Object-oriented database model with respect to building construction.
3.7. Cyber agents 4.1. Case study of control-recoÕery agent The term ‘‘agent’’ means someone acting on behalf of someone else. A computerized agent is an entity who has own mental state being composed of beliefs, capabilities, and commitment w7x. Beliefs
We executed a case study of agentification of control-recovery aids in the control, monitoring and warning computerized system for a semi-automated
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
sliding system of assembled roof within a underground power plant w8x. Fig. 4 shows schematic view of the prototype of control-recovery agents being developed in the case. The control, monitoring and warning computerized system here is functioning to: 1. Monitor the behavior of the assembled roof being slid, 2. Check whether its behavior is approaching its safety operating limits under operations, and 3. Automatically stop its behavior in emergency. The control-recovery agents translate the raw sensory data associated with the key variables into control or error signals, signs for displaying key warning and taking proactive actions, and symbols w6x for reasoning the underlying causes in order to
509
invoke error recovery functions at skill-, rule- and knowledge-based levels w6x. Signal is continuous quantitative indicator of the time-space behavior of the semi-automatic sliding system. The signal is categorized into control signal and error signal w6x. The control signal shows the transient state of behavior at a specific point in time. The error signal represents the differences between the actual state and the intended or planned state in a time-space behavior. The control-recovery agent stores a notion of the boundaries of acceptable performance Že.g., safety control limits. and the obligation to activate emergency shut down function if values of signal go beyond safety control limits in the capability. Workers can understand the existing state of the system by the control signal and off-normal state by the error signal. Signs are labeled by
Fig. 4. Schematic view of prototype of control-recovery agents.
510
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
names of states or situations of the semi-automatic sliding system w6x. Signs, which are built in IF–THEN rules, are used to select or activate stored predetermined actions that control the sequence of routine operations. Cyber-agent can display the key warnings in emergency and evoke built-in self-stopping and off-normal inspection functions when the raw sensory data goes beyond the control limit value. Workers can select and activate predetermined recovery actions by signs. Symbols refer to concepts tied to functional properties and can be used for reasoning and computation by means of a suitable representation of such property w6x. Since workers’ attention are likely to be directed at a small part of the total problem space at any one time, mimic diagram is relevant for them to direct their attention to logically important aspects of the problem space.
Information is treated as symbols in the mimic diagram to topographically investigate improper function and to symptomatically reason its underlying causes. The control-recovery agent here can denote the improper function topographically by making its location brightness and blinking in the mimic diagram. Workers can reason some potential faults based on symbols, as confirming and asking for additional data. 4.2. Case study on coffee-table type of construction method 4.2.1. Outline of Takada model Shimizu in Japan proposes a new construction method called coffee-table construction method for multi-stories apartment building. Fig. 5 shows
Fig. 5. Virtual dynamic construction with coffee-table method.
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
overview of building being constructed by this construction method. In this construction method, coffee-table modules are assembled in site factory and are transported and erected in order at the installation area. This method increases productivity and shortens construction duration compared with current construction method under stable conditions. Because we can assemble coffee-table modules under safety and invariable environment in site factory. In this case study, created is a virtual dynamic construction of the Takada model in the computerized test bed environment. 4.2.2. Outline of site factory As for division of main structure such as column, beam and load-bearing wall, details of the division
511
shall be standardized as considering conditions of design and construction method as a whole. The coffee-table construction method is composed of subsystems as for: 1. 2. 3. 4.
Transportation from supplier to site factory, Assembly works at site factory, Hoisting operation, and Operations of installation and connection. Job coordination is conducted by work rules among the subsystems and of themselves.
In this construction method, coffee-table modules are assembled at site factory and transported to a final installation position. At present, reasonable hoisting machine is tower crane. Size and weight of
Fig. 6. Virtual line simulation with coffee-table method.
512
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
the coffee-table module shall be designed as considering capacity of tower crane. Parts on truck are delivered from supplier to site factory. The parts are unloaded and stockpiled by hoist machine at unloading area. Hoist traveler grasps and transport the each part and select a rail to designated position according to the ID of part. As for subsystem at site factory, the following items shall be studied to develop optimal facility. Ø Number of rails for unloading and assembling, Ø Number of hoist traveler, Ø Precedence relationships between operations of machinery and equipment, Ø Conditions of entrance for trucks, Ø Assembly time for parts,
Ø Erection time for units, and Ø Bottleneck in process.
4.2.3. Production line simulation for coffee-table type of construction method This virtual dynamic construction centers on: 1. Multilevel assembly sequence ( Tooling operation ( Assembly operation ( Delay operation 2. Interference checks and Collision avoidance ( Part trajectory paths ( Swept volume envelope ( Dynamic cross-sectioning.
Fig. 7. Improvement of the computerized test bed environment.
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
513
Simulating the Takada model in the computerized test bed environment enables us to do research on:
calculated. The length of the time depends on each following item:
1. Planning: Determine work that can be released during certain time interval; 2. Scheduling: Determine work when that work is released into the field factory on a-minute-byminute basis; 3. Status and clock time: Know work position and machine status as well as ensuring that the plan representation covers the current clock time up to some predefined future plan horizon; and 4. What-if: ( Accept or reject the resulting plan; and ( Explore the consequence of changes to the current state of the field factory.
Ø The order of turning on PCa parts, Ø The time when PCa parts are turned on, Ø Layout of site factory and tower crane in construction site, Ø Transportation routes of PCa parts, Ø Ability of tower crane, and Ø Productivity of site factory.
Fig. 6 shows an example of line simulation of construction site where the coffee-table type of construction method is adapted. With this simulation, time necessary to build one layer of the building was
The combination of these items which satisfies a basic building plan can be obtained by repeating the line simulation with various conditions. We can improve the first construction plan and schedule referring to the result of the line simulation. 4.3. ImproÕement of the computerized test bed enÕironment The virtual construction simulation system will be able to create virtual construction in 3D computer-
Fig. 8. Configuration of VR-Coms.
514
Y. Maruyama et al.r Automation in Construction 9 (2000) 503–514
simulated environment before the planned construction is put into practice. The virtual construction simulation system is integrated with a planning, scheduling and performance management system, so that we will be able to find potential hazards latent in the plan being considered and to evaluate productivity. Fig. 7 shows mechanism to be built in the computerized test bed environment. Each task involved in construction works is grouped into the work packages, so that their forms can be identified in the 3D simulation and their progress can be monitored. The planning and scheduling systems are employed to produce the information regarding the work packages on the project and to determine when each work package is dispatched to the construction site. With the planning module and virtual simulation tools, ‘‘what-if’’ analyses can be carried out to explore what happens in the construction site, as well as when and where the events occur. The virtual simulation, when combining with the scheduling system, can be used to study the status of the project and to relate the work performed at a specific time and specific location. The control, monitoring and warning system monitors the behavior of the virtual construction simulation and captures the data needed to analyze productivity and safety. The coffee-table construction method being simulated in and this improvement of the computerized test bed environment will configure the VR-Coms as shown in Fig. 8.
5. Conclusion In this paper, the concept of VR-Coms was proposed and computerized test bed was built up. The computerized test bed of VR-Coms is integrated with virtual construction simulation, planning and scheduling, and real-time based construction management functions. First, we made the prototype of Control-recovery Agent on the test bed and attempted the operation confirmation of it. Second, we tried making the production plan of building work which used the coffee-table construction method. As a result, it has been understood that the test bed is useful as a
development environment of cyber-agent. We confirmed that the virtual construction simulation system integrated with planning, scheduling, and performance management function facilitate achieving interactive generate-and-test strategy to improve construction plans and schedules. Future work will concentrate on the configuration of VR-Coms improving the test bed environment. The anticipated results that this study incubates are summarized as follows: 1. Purchaser’s merits ( The content of the order is enhanced ( The design change is flexible ( Term of works shortening ( Reduction of construction expense 2. Cooperation trader’s merits ( An increase in unexpected construction expense is prevented ( The construction delay is prevented ( The arrangement to expect the big factor of safety becomes useless 3. Designer and contractor’s merits ( The design change is flexible ( Term of works shortening ( Reduction of construction expense.
References w1x S. Nishigaki, The application of discriminate analysis to safety patrol system, report of statistical application research, Union Jpn. Sci. Eng. 35 Ž1. Ž1988. 24–34. w2x S. Nishigaki et al., Humanware, human error, and Hiyari-hat: a template of unsafe symptoms, J. 1 of construction engineering and management, ASCE 120 Ž2. Ž1994. 421–442. w3x S. Nishigaki et al., Safety problems in on-site construction work processes, in: Proc. 11th ISARC, Brighten, U.K., 1997, pp. 13–18. w4x Construction Industry Institute Publication Nos: 3-1-Constructability: A Primer, July 1986. w5x Creation of building and urban city by intelligent system, Gihodo, 1998 ŽIn Japanese.. w6x J. Rasmussen, Skills, rules, and knowledge; signals, signs, and symbols, and other distinctions in human performance models, IEEE Trans. SMC SMC-13 Ž3. Ž1983. 257–266. w7x Y. Shoham, Agent-oriented programming, Artif. Intell. 60 Ž1993. 51–92. w8x S. Nishigaki et al., Control, monitoring and warning computerized system for a semi-automated sliding system of assembled roof, in: Proc. 14th ISARC,1997, pp. 199–205.