International Construction Specialty Conference of the Canadian Society for Civil Engineering (ICSC) (5th : 2015)

An automated model for selecting the optimum mobile crane model and on-site position using genetic algorithms Zaki, Tarek M.; Hosny, Osama; Nassar, Khaled Jun 30, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52660-Zaki_T_et_al_ICSC15_128_An_Automated_Model_For_Cranes.pdf [ 2.86MB ]
JSON: 52660-1.0076366.json
JSON-LD: 52660-1.0076366-ld.json
RDF/XML (Pretty): 52660-1.0076366-rdf.xml
RDF/JSON: 52660-1.0076366-rdf.json
Turtle: 52660-1.0076366-turtle.txt
N-Triples: 52660-1.0076366-rdf-ntriples.txt
Original Record: 52660-1.0076366-source.json
Full Text

Full Text

5th International/11th Construction Specialty Conference 5e International/11e Conférence spécialisée sur la construction    Vancouver, British Columbia June 8 to June 10, 2015 / 8 juin au 10 juin 2015   AN AUTOMATED MODEL FOR SELECTING THE OPTIMUM MOBILE CRANE MODEL AND ON-SITE POSITION USING GENETIC ALGORITHMS Tarek M. Zaki1,2, Osama Hosny1 and Khaled Nassar1 1 The American University in Cairo, Egypt 2 Abstract: Selecting both the optimum mobile crane model for a lift job and identifying collision-free position on site can result in productivity and safety improvements. Planning for lift operations is a task carried out by experienced lift engineers who examine each mobile crane’s lift-capacity specifications as provided by the cranes manufacturer to determine the most suitable mobile crane, configuration settings and a collision-free placing location. In this paper, an automated model was developed that follows a four-stepped algorithm process. First, is an algorithm that acquires the user defined lift requirements and calculates the required crane configuration settings using non-linear trigonometric equations. Second, is an algorithm that selects a crane model from the database of mobile crane models which mostly fits the calculated configuration in terms of crane safety and operation. Third, an algorithm that provides safety and clearance checks for the selected configuration. Finally, Genetic Algorithms are applied to optimize the process and to select the optimum crane model that has the safest configuration settings to use as well as its optimum placing location on site. The model was designed to generate a site layout plan as an output with a Cartesian coordinate system in order to assist the lift engineers in their planning for lift operations. The model was validated through its implementation on a working construction project in Cairo, Egypt. The outcomes highlighted the potential benefit of the model in assisting lift engineers in the planning of lift operations on construction sites and demonstrated its essential features. 1 INTRODUCTION Heavy lifts planning require the selection of an appropriate type of crane that is capable of carrying, slewing and placing a load without endangering the safety of the crane and the ongoing site operations. In general, mobile cranes are more suitable for typical infrequent heavy lifts such as placing mechanical equipment on building roofs, steel structure erection, etc. The task of selecting a mobile cranes and developing engineered-lift-studies is a crucial part in any heavy lift operation planning. This is due to the fact that, selecting the appropriate mobile crane for a specific lift contributes to the overall project’s productivity; thus, minimizing schedule delays and cost overruns (Hanna 1994). In practice, heavy lifts planning is prepared by experienced lift engineers. Their task is to (1) determine the most adequate crane configuration settings that can safely carry the required load and (2) identify the most appropriate placing location on site that is collision-free with all the nearby structures. A typical mobile crane’s specification has a number of charts including: model geometry and dimensions, lift-capacity charts and boom length range charts. Lift engineers use these charts to determine the best configuration in terms of: operating radius, operating boom angle to ground and the boom length extension which correspond to the weight of the object required to be lifted. Unfortunately, heavy lifts planning is a manual human-based calculation 128-1 which is prone to errors and requires some interpolation and guesswork between charts. Moreover, it does not guarantee the selection of the best mobile crane for the job and its most appropriate onsite location. However, the mobile crane selection and location problem is not new, previous studies have tackled this problem from different perspectives as will be discussed in the next section. 2 PREVIOUS STUDIES 2.1 Mobile Crane Selection A number of studies have focused on automating the mobile crane selection process by developing models and algorithms that determine the best configuration settings required to safely lift a given load.  Al-Hussein et al. (2001) developed a systematic approach algorithm for selecting mobile cranes and calculating the spatial clearances for their locations on construction sites. The work focused more on the development and processing of the algorithm for any type of mobile crane problem. The algorithm was coded with MS-Visual basic and supported by a previously developed crane database named “D-Crane” (Al-Hussein et al. 2000) which was designed to replace the lift-capacity charts for each crane model and rigging accessories. The working principle for the algorithm is that a crane data is retrieved from D-Crane and is processed to determine the near optimum selection of configuration settings which are showed on a user interface.  Moselhi et al. (2004) developed a system that integrates the previously developed algorithms (Al-Hussein et al. 2000 and 2001) with graphics engine to produce 3D animations that act as a simulation tool for a lift planning task. The system demonstrated its capabilities as it assisted crane operators in avoiding potential accidents and reduced the time and cost associated with planning the lift operation on site.  Later in 2005, Al-Hussein et al. developed another algorithm which was not restricted to predefined configurations of cranes’ manufacturers, the objective was to eliminate the guesswork required by practitioners in relation to the limited information pertained in the lift-capacity charts. The developed algorithm continued the work of the previously developed algorithm (Al-Hussein et al. 2001) and provided the optimum crane configuration for a required lift as a range of minimum and maximum values of boom length and radii settings. The algorithm was also developed with MS-Visual basic and the optimization module used MS-Solver. The output of the algorithm was displayed as a 2D-elevation for the range of crane settings.  Wu et al. (2010) developed an algorithm for selecting mobile cranes that considers the lift capacity, geometric characteristics of the crane and the ground bearing pressure which was one of (Al-Hussein et al. 2005) limitations. The algorithm was incorporated in a 3D environment to simulate the crane operation. The crane selection module only covered crawler crane types. The path planning was not considered in the 3D simulation. The proposed algorithm was applied on a working construction project which demonstrated the usability and accuracy of the developed algorithm. 2.2 Workspace planning for the onsite locations of mobile cranes Most of the previous studies discussed above were mainly concerned with the development of algorithms that automate the selection of the optimum mobile crane configuration settings. However, the collision free-paths and a crane’s workspace considerations in relation to the site plan was not tackled enough.  Tantisevi and Akinci (2009) have modeled the dynamic behavior of mobile cranes during operation in order to identify the possible spatial conflicts. The developed model generated the motion of mobile cranes during construction given some inputs including: building design, schedule, crane load-capacity specifications and information regarding the crane operation. The output of the model generated a 4D simulation of mobile cranes on construction sites that identified the possible spatial conflicts and determined the set of conflict-free site locations where a crane can be placed.  Safouhi et al. (2011) developed an algorithm to determine the optimal location for mobile cranes onsite using a mobile crane’s body working area by defining inside and outside boundary limits as restriction 128-2 areas for a crane’s operation. The outside boundary limit was defined as the access areas around permanent structures, while the inside boundary limit was defined as the areas which pertain the crane’s operation due to obstructions such as laydown areas, construction trailers, foundations, etc. The intersection between the inner boundary limit and the outer boundary limit was defined as the feasible area for the crane’s operation. The model used a 2D construction layout plan where facilities and the mobile crane are presented as (x,y) Cartesian coordinates. The algorithm was validated via a case study; the output generated all the feasible locations for a crane to be placed with the ability to rotate 360˚ without its body colliding with any onsite obstruction. After reviewing the previous works, some findings could be drawn: (1) the previously developed crane selection algorithms used mathematical optimization techniques which adopt differential equations to determine the maximum and minimum values, (2) the mobile cranes’ databases were mostly developed while the original specification charts supplied with each mobile crane model were not considered, (3) the crane’s workspace planning required simulation in a 3D environment to determine the spatial conflicts with the surrounding obstructions.  The objectives of this paper are to develop an automated model that uses the original lift-capacity charts supplied with the mobile cranes, optimize the selection, location and configuration settings of the mobile crane using Genetic Algorithms (GA) while considering the current site mobilization plan. 3 METHODOLOGY A model was developed that is composed of a three components: Inputs, Engine and Outputs as shown in Figure 1. The Inputs component is a site layout plan module which includes coordinates for the site boundaries, the permanent facilities, the temporary facilities and the object required to be lifted. The Engine component, which is the core of the model, is composed of three concurrently working modules: cranes database, geometry calculator and safety/clearance checks. Each module has its own built-in calculating algorithm and all modules are linked together. GA was used to determine the optimum output from each of the three modules while satisfying a set of constraints. Finally, the Output component generates the most suitable crane to be used, its configuration and the best onsite operating location that is conflict-free. The output is presented on a site layout plan. The developed model was coded on MS-Excel 2010 and was designed with a user interface to facilitate the user inputs and the output reports. GA was used in this model due to the combinatorial nature of the problem in hand. Moreover, GA tend to find  global optima in complex spaces compared to traditional mathematical optimization, as presented in the  previously developed models that may get stuck in local optima (David 1991).  The working principle of the proposed model is that, at first the site mobilization data which includes the site boundaries, temporary facilities and object to be lifted is input to the model in the form of Cartesian coordinates. At the same time, the Crane Database module generates a crane with all its specifications as placing coordinates as an initial solution to the GA engine. Second, the Geometry Calculator module determines the boom length and working radius that correspond to the required weight to be lifted using a set of non-linear trigonometric equations. The calculated length and radius then flow back to the selected crane’s lift-capacity charts following a search algorithm to select a built-in configuration of boom length, working radius and maximum weight for this setting. The selected configuration then passes to the Safety & Clearance Checks module where a checks algorithm determines any flaws with the selected configuration. If any of the check tests fail, the crane database module selects another crane model and new coordinate points to be the new initial solution for the model and starts over again. If the configuration passes all the check tests, an output report is generated which includes the selected crane model, its configuration, it’s placing location on site considering the previously input mobilization plan. 3.1 Module 1: Site Layout Plan The site layout plan was modeled in 3-D space where all the facilities, the mobile crane and the object to be lifted were represented in the form of blocks, each defined by nine Cartesian (x,y,z) coordinates as shown in Figure 2.  128-3 Where “2**” in Equation 13 is an extra 1 meter allowance from each side of a facility and the crane as a common practice recommendation. 3.4.4 Boom to building collision The operating angle should be sufficient enough to avoid any collisions with the building when planning a lift; therefore, this can be depicted using Equation 14 which requires that the clearance should be bigger than the maximum of either the centroid coordinates location of the weight to the edge of the building. In common practice, usually the minimum clearance is 1 meter from any of the building edges. [14] (Hr + Hw) / Tan Ɵ > Max (Abs(x2(building) – xC(weight)) OR Abs(y2 (building) – weight yC (weight))) 3.4.5 Weight to building intersection The hook of the crane in the rigging element cannot intersect with the building as shown in Equation 15. [15] L Sin Ɵ ≥ (Hb – P2G) + Hr + Hw 3.4.6 Boom to Weight Clearance The rigging height has to be tall enough to avoid boom collision with the weight as shown in Equation 16. [16] Hr / Tan Ɵ ≥ b(weight)/2 4  CASE STUDY The case considered herewith involved the placement of an Air Handling Unit (AHU) on top of an office building project located in Cairo, Egypt. The building was a 6-story high; the mobilization plan for the site was obtained from the Contractor and was input to the developed model. The site mobilization plan was composed of a number of facilities; including: Employer/Engineer’s offices, Contractor’s offices, storage/laydown areas, carpentry workshop and steel workshop. The Contractor’s fleet of owned mobile cranes were also considered and included a number of 4 all-terrain telescopic boom mobile cranes of different capacities. The site layout data was input through the model interface which is divided into 3 sections as shown in Figure 7. The first section is about the permanent facilities which require the user to define site boundaries (length and width), the dimensions and coordinates of the buildings being constructed. The model allows for three buildings to be placed in the layout; however, this case only uses one building; thus, the other input spaces were not used and are shown to be dimmed on the model interface. The second section is about the temporary facilities. The model allows the user to enter 6 temporary facilities; however, in this case only 5 were used. The last section is about the object to be lifted were the user is required to define the object’s dimension, coordinates and the weight in tons, which in this case is the AHU. The objective function was originally set to minimize the radius while satisfying the set of constraints as defined in the model engine. The GA initial solutions’ pool was set to be 100 solutions, the stopping time was set to 15 minutes, the crossover rate was set to 95% and the mutation rate was set to 5%. The model was required to run a number of times to validate its outputs. 5 RESULTS Evolver™ was left to complete its 15 minutes run time, even though the solutions showed to have stopped converging after 10 minutes as shown in Figure 8. The algorithm selected crane model “DEMAG AC (665) 250t” to be used with the configuration of L= 53.6m that corresponds to R= 34m, Ɵ=55.5˚, the placing coordinates (43.4,4.6), the outriggers to be fully extended, the counter weight= 100t and the operating zones to be 360˚. The outputs were presented in the form plan view, east elevation and south elevation, whereas the calculate parameters were represented on the mobile crane schematic diagram as shown in Figure 9.  By further running the model extra number of times, it was concluded that for this optimum configuration, other two optimum placing coordinates could be used, which are: (75.3,27.8) or (60,4.7). By these results, the Contractor had the flexibility to select the onsite placing location for the 128-8 mobile crane from these 3 optimum locations (coordinates) in order to minimize any disruptions to the ongoing construction activities that may arise due to the mobile crane’s placing position and operation.  Figure 7: Model interface with inputs  Figure 8: GA optimization outcome after 10 minutes 6 LIMITATIONS AND FUTURE IMPROVEMENTS Although the proposed model demonstrated practicality, yet it had some limitations including: (1) the site layout module was designed to include only 6 temporary facilities and 3 permanent facilities. However, such limitation can be adjusted from the model engine itself, (2) the mobile crane database only considered the telescopic boom cranes without the allowance of an additional jib attachment, and (3) the mobile crane was considered to be a block diagram and the boom a straight line, more detailed variables may need to be included in the model calculations such as the soil bearing pressure, the wind load, etc. 7 CONCLUSION The task of selecting a mobile cranes and developing engineered-lift-studies is a crucial part in any heavy lift operation planning. In practice, experienced lift engineers develop such studies manually and interpolate between the specifications charts to select the mobile crane to uses while estimating the corresponding onsite position. A model was developed to automate this process and provide the optimum crane selection and onsite position. The model was coded on MS-Excel and ran with GA using Evolver® add-in for excel. A case study for an ongoing construction project in Cairo, Egypt was presented to demonstrate the use of the developed model and illustrate its features. The outcomes demonstrated the 1. Permenant Facilities 2. Temporary Facilities 3. Object to be LiftedWeight13 tonL W L W H L W H L W H L W H120 85 65 55 18 20 15 3 30 15 3 4 3 2X1 Y1 X1 Y1 X1 Y1 X1 Y15 25 104 5 104 27 50 35L W H L W H L W H30 20 3 10 20 3X1 Y1 X1 Y1 X1 Y189 63 2 2L W H L W H L W H10 20 3X1 Y1 X1 Y1 X1 Y12 13L I F T   P L A N N E RI. Site Layout PlanAHU - 01Carpentry WorkshopSteel WorkshopEmp/Eng Offices Contractor OfficesStorage and LaydownSITE BuildingC A L C U L A T E S T O P(x1,y1)xyLW128-9 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items