Integration of Ergonomic Analysis into Simulation Modelling of Manual Operations

Transcription

1 Simulation in Production and Logistics 2015 Markus Rabe & Uwe Clausen (eds.) Fraunhofer IRB Verlag, Stuttgart 2015 Integration of Ergonomic Analysis into Simulation Modelling of Manual Operations Integration der ergonomischen Analyse in die Simulationsmodellierung von manuellen Operationen Alireza Golabchi, SangUk Han, Simaan AbouRizk, University of Alberta, Edmonton (Canada), Abstract: Considering the prevalence of simulation modelling in production design, integrating ergonomic safety considerations into simulation models can help production planners identify and prevent ergonomic risks before workers encounter unsafe conditions, in addition to productivity improvements. This study proposes integrating ergonomic analysis into micro-motion-level simulation modelling by developing a simulation modelling template that enables simultaneous estimation of the standard duration of manual activities, as well as biomechanical risk factors inherent in the production design. The proposed approach enables modelling of manual activities at a human motion level in order to provide standard task durations in addition to initial insight into the level of ergonomic risks associated with a planned production scenario. As a case study, the proposed methodology has been implemented in a construction steel fabrication company and the results confirm the validity of the approach in identifying ergonomic risks associated with a production design. 1 Introduction Simulation modelling is an effective tool for design and planning of construction operations. A digitalised model representing a physical system allows prediction of the system performance in a real world application, as well as understanding of the system behaviour under various conditions. In construction, advances in simulation modelling techniques have been made to estimate the duration of projects and the productivity of different operational systems. However, it still remains challenging to analyse manual operations due to the uncertainties in modelling human movements and estimating the cycle time of manual tasks. Furthermore, the impact of the designed workplace and workflow on human capacity (e.g., durations and frequencies of physical tasks) for ergonomic safety analysis has not been thoroughly considered. In this regard, integrating ergonomic considerations into simulation models can help production planners identify and prevent ergonomic risk factors

2 492 Goolabchi, Alirezza; Han, SanggUk; Abourizkk, Simaan affectingg human moveements beforee the workers encounter thee unsafe condditions, as well as understand u thee impact of prooduction systeems on humann behaviour. To addreess this issue, this study prooposes integraating ergonom mic analysis innto micromotion-llevel simulatioon modelling. A discrete-eevent simulation modelling template has beenn developed which w enables simultaneouss estimation of the standardd duration of manuual activities and biomechhanical risk factors inhereent in the prroduction design. Specifically, S t concept of a Predeterm the mined Motion Time System m (PMTS) is adopteed to model human h movem ments (e.g., rettrieving, carryying, placing aan object) and estim mate cycle tiime of manuual tasks baseed on the unnit cycle timee of each movemeent. The humaan postures annd motions in nvolved in thee manual operration are then evaaluated from an a ergonomic perspective, using u an ergonnomic assessm ment tool. The propposed approacch thus provides the user with w initial innsight into thee level of ergonom mic risks assocciated with a planned prod duction scenarrio. By inputtting basic design data d available during workpplace design, the simulationn model will report on the ergonnomic risks asssociated withh the task baseed on posturall strains, exterrnal force exertion,, and frequenncy of the task. As a resu ult, the simullation model not only providess information regarding thee standard du uration of the envisaged prroduction, but also reports on thhe level of erggonomic riskss associated with w the desiggn; which potentiallly allows for f the in-deepth understaanding of thhe trade-off between productivvity and safe fety. Consequuently, the prroposed simuulation modell enables modifyinng the workplace design inn order to acchieve both errgonomically safe and productivve operation systems. s 2 Method This research uses PM MTS and erggonomic assesssment in ordder to developp an integrated micro-motion-l m level simulatioon modelling environment. The framewoork of the research is shown in Figure F 1. Figure 1: 1 Integrated micro-motionm -level simulatiion framework rk 2.1 Predeterm mined Motio on Time Sy ystem PMTSs are used in various v industrries to determ mine the amouunt of time reequired to perform a manual proocess by breakking down a task t into sequuential motionns carried out by thhe worker to complete c the task t (Farrell 1993). 1 By usinng PMTSs, prroduction planners can make moore productive decisions on n designing work w processes, as they

3 Integration of Ergonomic Analysis into Simulation of Manual Operations 493 are provided with a standard duration in which a worker is expected to complete an action, as well as a comparison of the efficiency of different alternatives of performing the operations (ILO 1986). Researchers have studied large samples of various manual activities in order to develop PMTS methods that can provide the standard time required to carry out a manual task. Currently, the most widely used PMTSs include Methods-Time Measurement (MTM) (Maynard et al. 1948), Modular Arrangement of Predetermined Time Standards (MODAPTS) (Heyde 1966), and Maynard Operation Sequence Technique (MOST) (Zandin 1980). This study uses the MODAPTS method for modelling worker motions due to its simplicity and quick application. However, the other methods can also be similarly applied using the same approach. MODAPTS classifies the complexity of manual actions (e.g., get, move, put) by the amount of sensory feedback required to carry out the action rather than the geometrical properties of the material to be handled. It was developed in an attempt to reduce the time and cost required to apply prior PMTS methods and create a system that could be memorised, is simple to learn and apply and yields consistent results (Heyde 1989). MODAPTS assumes that the duration of any body motion can be expressed as a multiple of the time required to move a single finger, called a MOD. The MOD unit, equal to second, is used assuming that the movements are carried out with minimal energy expenditure and that the movement duration is proportional to the fifth root of the moment of inertia of the body part moved (Minter 1983). Figure 2 shows the moving distance of the different body parts and the corresponding MOD values for MODAPTS. Body Part Finger Hand Arm Whole arm Extended arm Trunk MOD Distance Moved 2.5 cm 5.0 cm 15 cm 30 cm 45 cm 75 cm Figure 2: Basic movements as defined by MODAPTS In order to analyse a manual task using MODAPTS, the task has to be broken down into basic motions, known as modules, that can be described using MODAPTS classes, and for each motion, a MOD value has to be assigned. By adding the MOD values and converting the sum of MODs to seconds, the total amount of time required to complete the task is obtained. MODAPTS codes are used to represent different classes of movements which describe the type of motion. Some basic

4 494 Golabchi, Alireza; Han, SangUk; Abourizk, Simaan classes include: Get (G), Move (M), Put (P) and Use (U). For example, a worker moving his hand 15 centimetres to pick up a wrench is represented by the MODAPTS code M3G3. M3 represents the hand movement motion that takes 3 MODs, and G3 represents performing an impeded grasp which also takes 3 MODs. For the purpose of this study, the MODAPTS method is integrated into discreteevent simulation in order to enable the modelling of manual activities in simulation models of construction operations. 2.2 Ergonomic Assessment Various ergonomic assessment methods have been developed in order to evaluate the risks associated with manual tasks and identify unsafe actions that can lead to Work-related Musculoskeletal Disorders (WMSDs). These methods consider human posture as well as external risk factors (e.g., frequency and duration) to provide a global risk assessment of a motion, which is used to take required corrective actions (Golabchi et al. 2015b). Different methods prioritise different factors and focus on various body parts in order to assess different cases of manual handling activities. The cube model (Kadefors 1994) is one of these ergonomic assessment tools which provides an ergonomic evaluation of a manual task by inputting basic design data. This study uses the cube model for ergonomic evaluation due to its simplicity and efficiency, which make it suitable for incorporating into simulation modelling. The cube model was developed based on the premise that the risk of acquiring WMSDs is mainly dependent on three interrelated variables (Laring et al. 2002). These variables include: working posture, muscular force, and time. For each of these factors, three levels of demands are defined based on severity, including low demand, medium demand, and high demand. These demand levels can be assigned using scientific evidence (e.g., Snook and Ciriello s (1991) tables for maximum acceptable weight limits), or by using a consensus approach to identify the severity of each demand (Kadefors 1993). The three factors (i.e., posture, force and time) are considered as axes of a cube, yielding 27 sub cubes, which is the number of possible combinations. For each combination, a score is calculated by multiplying the scores of each of the three demands, assuming 1, 2, and 3 as low, medium, and high demand, respectively (Kadefors 1997). A manual task is considered ergonomically acceptable for a final score of less than 5, conditionally acceptable for a score between 5 and 10, and ergonomically unacceptable for a score of higher than Integrated Micro Motion Level Simulation The authors use Simphony (Hajjar and AbouRizk 1999) as the platform for simulation modelling as it enables development of special purpose templates that can encompass elements that work in conjunction with the general template. For the purpose of this study, an integrated micro-motion-level simulation template is developed that consists of elements that enable applying the MODAPTS standard in the simulation environment, without requiring knowledge about the details of implementing MODAPTS, and also provide insight into the level of ergonomic risks associated with manual tasks of the model, based on the principles of the cube model. The modelling elements developed include one primary element, named MHE (Manual Handling and Ergonomics), and six secondary elements. The MHE element requires basic design data pertaining to a manual task as input and

5 Integration of Ergonomic Analysis into Simulation of Manual Operations 495 calculates the corresponding MODAPTS duration for the task and uses it for simulation purposes, and also reports on the level of ergonomic risks for each manual task. Figure 3 shows the design inputs required for the MHE element. As shown in Figure 3, the required inputs include design information that a production planner has when designing or redesigning a process. Inputs required for MODAPTS calculation Inputs required for ergonomic assessment Figure 3: Inputs of the MHE element The six secondary elements represent lower level manual activities that are designed based on the MODAPTS classes and include: Move, Get, Put, Walk, BendAndArise and SitAndStand. For example, the Move element can represent the motion of a worker moving his hand to grasp or put down an object, with the required input being the distance between the worker s hand and the object. These elements are designed and implemented to provide more flexibility in designing detail-level manual motions, in cases where a designer intends to investigate the effect of changing the attributes of one motion on the simulation results (Golabchi et al. 2015c). Furthermore, these models are used when the model requires placing one lower-level motion between non-manual elements. Figure 4 shows the Simphony environment and the elements of the special purpose template. By using the developed elements, designers will have the opportunity to use reliable estimates of the duration required to perform a manual task within the simulation environment, without requiring prior knowledge about the details of predetermined motion time systems. Furthermore, designers are provided with an assessment of the level of ergonomic risks associated with each manual activity, which enables them to prevent WMSDs during the workplace design phase.

6 496 Golabchi, Alireza; Han, SangUk; Abourizk, Simaan Figure 4: Simphony modelling environment 3 Implementation: Case Study In order to implement and validate the proposed approach, the integrated micro motion level simulation has been used to model a steel plate handling task in a steel fabrication and construction service provider company in Canada (Golabchi et al. 2015c). Due to the nature of manual tasks in the construction industry, workers are frequently involved in physically challenging activities. This makes the integrated simulation approach suitable for modelling these manual tasks since both a standard duration for worker actions and a feedback on the level of ergonomic risks associated with the action is obtained. The steel plate handling task investigated as the case study includes a worker removing a steel plate from a drilling machine after the drilling is done, carrying the plate to designated bins, and placing it in the appropriate spot inside the bin. This cycle can be modelled for simulation purposes using the developed MHE element. The drilled plates have various dimensions and weights with an average size of 40 cm x 40 cm x 5 cm and an average weight of 20 kilogrammes. The task has been observed at the jobsite in order to collect the required inputs for the MHE element. Table 1 shows the inputs required and the description of each input for modelling the plate handling task. Although the proposed approach can be highly effective in designing new processes, an existing manual task has been selected for the case study in order to firstly enable studying the validity of the approach and secondly demonstrate its functionality in redesigning existing operations, besides designing new processes. It should be noted that the MHE element modelling the steel plate handling task will be used as part of a larger simulation model in conjunction with other simulation modelling elements, representing the whole steel fabrication operation. The steel plate handling activity can also be broken down into lower level motions and be modelled using the secondary elements.

7 Integration of Ergonomic Analysis into Simulation of Manual Operations 497 Table 1: Inputs required for modelling the steel plate handling task Input Description for steel plate handling Alternatives Distance EndPosition The distance that the worker carries the plate. The precision required for placing the plate at the destination. - GeneralLocation, WithTidiness, ExactLocation Frequency The frequency of the task per day. LessThanOneHour, BetweenOneAndFour, MoreThanFourHours Posture RetrievalEnd/ RetrievalStart The optimality of the worker s posture while handling the plate. The distance between the plate and the worker s hands when retrieving/placing the plate. Optimal, NearOptimal, OutsideOptimal OneInch, TwoInches, SixInches, TwelveInches, EighteenInches, ThirtyInches StartGrasp The ease of grasping the plate. SimpleGrasp, ImpededGrasp Weight The weight of the plate. - Validating the proposed approach using the case study involves two steps: (1) comparing the MODAPTS duration calculated by the model with actual time data collected from the jobsite, and (2) investigating the reliability of the ergonomic assessment provided by the simulation model. In order to perform the first part of the validation study, the actual duration for the worker to complete the task is also recorded for each instance, besides collecting the data related to the inputs shown in Table 1. Table 2 shows the results of running the simulation model using the MHE element for 10 instances of the plate handling task as an example. The MODAPTS code and duration are also calculated manually to ensure that the simulation model is calculating the durations accurately, and the results are consistent with the durations calculated by the simulation model, as shown in Table 2. A correlation analysis has been carried out and a Pearson correlation coefficient of and a Spearman correlation coefficient of is calculated between the actual time and the simulation MODAPTS time, which indicates very strong association between the two data sets. In order to investigate the reliability of the ergonomic assessment results from the developed simulation template, for each instance of the steel plate handling task, the level of ergonomic risks associated with the instance is analysed using the Rapid Upper Limb Assessment (RULA) (McAtamney and Corlett 1993) method. RULA is a widely-used approach to ergonomic assessment which focuses on human posture and also considers the load involved as well as the frequency of the task (Golabchi et al. 2015a). The result of a RULA analysis is a total score between 1 and 7, with 1 and 2 indicating acceptable posture, 3 and 4 indicating that changes may be needed, 5 and 6 indicating that changes are required soon, and 7 indicating that immediate

8 498 Golabchi, Alireza; Han, SangUk; Abourizk, Simaan investigation and modification is required. Since the result of a cube model analysis is represented in one of the three assigned categories (i.e., acceptable, conditionally acceptable, and unacceptable), the RULA scores are also categorised as 1 and 2 representing safe motion, 3 and 4 indicating further investigation required, and 5, 6 and 7 specifying unsafe motion, to enable comparison of the results. Table 3 shows the result of the ergonomic analysis for 10 instances of the steel plate handling task. The results indicate consistency between the output of the ergonomic assessment from the simulation model and the RULA analysis, which confirms the reliability of the approach in evaluating the ergonomic risks of manual activities. Table 2: Comparison between actual time and simulation time for case study Instance Duration (Sec) MODAPTS Code MODAPTS Actual Simulation (Manual) Duration (Manual) M2G1L3W47M2P (=55*0.129) M3G3L3W39M4P (=54*0.129) M3G3L3W35M4P (=50*0.129) M2G1L3W39M5P (=50*0.129) M2G1L3W35M5P (=46*0.129) M3G1L3W35M5P (=47*0.129) M3G1L3W35M5P (=49*0.129) M3G3L3W23M2P (=34*0.129) M2G3L3W31M1P (=42*0.129) M3G3L3W31M2P (=44*0.129) Table 3: Comparison between results of ergonomic assessment and RULA Instance Cube model result RULA Score Interpretation 1 Conditionally acceptable 3 Further investigation 2 Unacceptable 6 Unsafe 3 Unacceptable 6 Unsafe 4 Conditionally acceptable 3 Further investigation 5 Unacceptable 6 Unsafe 6 Unacceptable 6 Unsafe 7 Conditionally acceptable 3 Further investigation 8 Conditionally acceptable 3 Further investigation 9 Unacceptable 6 Unsafe 10 Unacceptable 6 Unsafe

9 Integration of Ergonomic Analysis into Simulation of Manual Operations 499 The results of the case study analysis indicate that the integrated micro motion level simulation approach can be effectively used by process designers and production planners to provide firstly a standard duration for manual tasks and secondly an insight into the level of ergonomic risks associated with a design. The correlation between the actual and simulation time can be used as a benchmark for worker s performance which enables evaluating the efficiency of operations for work process design and redesign purposes. In future work, the authors will consider linking human motion data to the elements of the developed special purpose simulation template in order to enable more accurate ergonomic evaluation and mitigation of ergonomic risks while redesigning ongoing operations. Furthermore, more data from jobsites will be collected in order to formulate the conversion of standard MODAPTS time to realistic jobsite time by accounting for inefficiencies in carrying out manual tasks. The result can be used to develop efficiency factors for various types of manual activities. 4 Conclusion Simulation modelling is an effective tool for design and planning of operations since a digitalised model representing a physical system allows for prediction of the system performance in the real world, as well as for understanding system behaviour under various conditions. In construction, advances in simulation modelling techniques have been made to estimate the duration of different types of projects and the productivity of various operational systems. However, it still remains challenging to analyse manual operations due to the uncertainties in modelling human movements and estimating the cycle time of manual tasks. Furthermore, although production planners intend to take ergonomic considerations into account while planning operations, there is a lack of tools to enable ergonomic safety assessment during workplace design. In this regard, this study proposes integrating ergonomic analysis into micro-motion-level simulation modelling, which enables simultaneous estimation of the standard duration of manual activities and ergonomic risks inherent in production design. A special purpose simulation template is developed that enables modifying a workplace design in order to achieve both ergonomically safe and productive operation systems and potentially allows for the in-depth understanding of the trade-off between productivity and safety. This study provides industry practitioners with the opportunity to incorporate reliable representations of manual activities into simulation models of operations and take advantage of standard PMTS methods as well as ergonomic assessment tools, without requiring prior knowledge about these systems. References Farrell, J.M.: Predetermined Motion-Time Standards in Rehabilitation, A Review. Work: A Journal of Prevention, Assessment and Rehabilitation 3.2 (1993), pp Golabchi, A.; Han, S.; Fayek, A. Robinson (2015a) A Fuzzy Logic approach to posture-based ergonomic analysis for field observation and assessment of construction manual operations. Canadian Journal of Civil Engineering (2015). (in review, submitted March 2015)

10 500 Golabchi, Alireza; Han, SangUk; Abourizk, Simaan Golabchi, A.; Han, S.; Seo, J.; Han, S.; Lee, S.; and Al-Hussein, M. (2015b) An automated biomechanical simulation approach to ergonomic job analysis for workplace design. Journal of Construction Engineering and Management 141 (2015) 8, online. Golabchi, A.; Han, S.; AbouRizk, S.M. (2015c) Integration of predetermined motion time systems into simulation modelling of manual construction operations. In: Proceedings of the 5th International/11th Construction Specialty Conference, Vancouver (Canada), June (accepted for publication) Hajjar, D.; AbouRizk. S.: Simphony: An environment for building special purpose construction simulation tools. Proceeedings of the 1999 Winter Simulation Conference. Phoenix (AZ) 1999, pp Heyde, G.C.: Modapts. Industrial Engineering (1966), pp Heyde, G.C.: Concepts and history of ModaptsPlus, Industrial Engineering 30 (1989), pp ILO: Introduction to work study (3rd Edition). Geneva: International Labour Organization (ILO) Kadefors, R.: A model for assessment and design of workplaces for manual welding. In: Marras, W.S., Karwowski, W., Pacholski, L. (eds.): The Ergonomics of Manual Work. London: Taylor & Francis Kadefors, R.: An ergonomic model for workplace assessment. In: Proceedings of the IEA 94, Vol. 5. International Ergonomics Association, Toronto (Canada) 1994, pp Kadefors, R.: Evaluation of working situations using the cube model approach. In: Proceedings of the IEA 97, Vol. 4. International Ergonomics Association, Tampere (Finland) 1997, pp Laring, J.; Forsman, M.; Kadefors, R.; Örtengren, R.: MTM-based ergonomic workload analysis. International Journal of Industrial ergonomics 30 (2002) 3, pp Maynard, H.B.; Stegemerton, G.J.; Schwab, J.L.: Methods-time measurement. Boston: McGraw-Hill McAtamney, L.; and Corlett, E.N.: RULA: A survey method for the investigation of work-related upper limb disorders. Applied Ergonomics 24 (1993) 2, pp Minter, A.L.: Modapts. Management Services 27 (1983), pp Snook, S.H.; Ciriello, V.M.: The design of manual handling tasks: Revised tables of maximum acceptable weights and forces. Ergonomics 34 (1991) 9, pp Zandin. K.: MOST. Work Measurement Systems. New York: Dekker 1980.