Impact of selected work condition factors on ... - Wiley Online Library

Impact of Selected Work Condition Factors on Quality of Manual Assembly Process ´ Adam Hamrol, Dagmara Kowalik, and Agnieszka Kujawinska Poznan´ University of Technology, Institute of Mechanical Technology, Poznan, ´ Poland

Abstract This article presents the investigation of selected work environment factors with regard to the quality of manual assembly process. It describes the research carried out in a natural work environment and conditions in a factory producing car wire harnesses. It has been pointed out that only a few work condition factors have a significant influence on the assembly process’s quality. The results of the experiments have proved the hypothesis that, between the two important work factors (noise level and work monotony), there is a significant interaction with respect to their impact on the assembly quality process. C 2010 Wiley Periodicals, Inc. Keywords: Assembly process; Work condition factors; Quality of the process

1. INFLUENCE OF ENVIRONMENTAL FACTORS ON THE HUMAN IN MANUFACTURING PROCESSES Despite all the recent advances in automation and the introduction of information technology into manufacturing processes, they still depend on human participation and intelligence. In some kind of processes, human participation is even prevailing. First of all, it refers to the manual assembly processes with simple and repeatable work operations performed by a human. The role of humans in manufacturing processes can be both positive and negative. On the one hand, with regard to the efficiency and quality of the production processes, human participation has remarkable advantages: A human can be creative and able to adapt to unexpected conditions and solve uncommon problems. ´ Correspondence to: Agnieszka Kujawinska, Poznan´ University of Technology, Institute of Mechanical Technology, Piotrowo 3, Poznan´ 61-135, Poland. Phone: +48616652798; e-mail: [email protected] Received: 22 July 2009; revised 25 November 2009, 3 February 2010; accepted 24 March 2010 View this article online at wileyonlinelibrary.com. DOI: 10.1002/hfm.20233

156

On the other hand, the human is the most fallible element in the production line, due to the limited mental and physical endurance that sometimes cause behavior and reactions to be unpredictable. In other words, the human has an inborn inclination to failures, but at the same time the human can realize his or her faults ´ and their reasons. According to Wasinska (1999) and Karwowski and Salvendy (1994), the human can be referred to as “the main process regulator in emergency situations.” The human involved in a production process is subject to various factors. In this article, the factors related to the human work conditions called Work Factors (WF) have been analyzed. Their influence may be positive, negative, or neutral. The factors can be divided into two groups according to the concept presented in Figure 1: •

•

Material environment factors: vibration, noise, microclimate (temperature, humidity and air movements, thermal radiation), lighting, emission of harmful energy (electromagnetic and electrostatic field), air pollution (Szulczyk & Cempel, 2006); techno-organization factors: body position, work rhythm and pace, intervals (breaks for

Human Factors and Ergonomics in Manufacturing & Service Industries 21 (2) 156–163 (2011)

c 2010 Wiley Periodicals, Inc.

´ Hamrol, Kowalik, and Kujawinska

Impact of Selected Work Condition Factors

Material Environment Factors Mechanical vibrations

Body position

TOOL

Microclimate

MACHINE

Work subject

Noise

Work rhythm and pace

Lighting

Radiation

Signal unit

Receptors

Control unit

Efectors

Intervals

Work method

Pollution

Knowledge Memmory Gifts MAN Intelligence Motivation Skills

Manuals

Supervision

Techno – organizational Factors

Figure 1

Work factors in the system “Man-Machine-Environment” (Kowalik, 2008).

rest), production methods, manuals, data, supervision (Cheng & Jiang, 1995; Fitts, 1951; ´ Springer & Gorny, 2008; Stanton et al., 2004; ˙ Tytyk, 2008; Zurek et al. 2007). There are many articles concerning investigations of the impact of WF on human work efficiency and quality in manufacturing processes. These articles are concerned with relationships between a human and the technological object (Bradley & Hendrick, 1994; Broberg, 2007; Helander, 2006; Mantura, 2008; Senge, 2008; Singleton, 1974; Singleton et al., 1971; Tytyk, 2004; Wilson et al., 1987). Most of them are devoted to WF only; they do not deal with the interaction among the WF. It is obvious, however, that interaction is important from the point of view of optimizing the human work conditions and making the work environment friendlier.

wire harnesses are shifted from one work station to another, and the production pace is determined by the pace of shifting of the assembly line (Figure 2). It should be emphasized that the quality of the wire harness assembly process is equivalent to the quality of a human’s work because human failures are practically the only ones that can occur in this kind of assembly process (Karwowski & Salvendy, 1994). QAP was measured, in the investigation presented later in text, by an index ppminsp (ppm = parts per million). The index ppm determines the number of failures detected in the control process per one million produced car wire harnesses. These failures include lack of electrical contact, short-circuit, unscrewed cable, damaged housing, and so forth. This index is a sum of two others indices (see Figure 3) describing the number of failures detected by the fitters still on the

2. RESEARCH SUBJECT Worker 11

Worker 8

Worker 9

Worker 10

Worker 5

Figure 2

Worker 4

Worker 3

Worker 1

Worker 6

Worker 7

Assembly stand

The investigation of the selected work environment factors with respect to their impact on the quality of the car wire harness assembly process (quality of the assembly process [QAP]) is presented in this article. The car wire harness is a bundle of isolated cables with electrical contacts used in car electrical circuits. In the assembly process of this kind of product, work conditions are extremely hard. The workers (the fitters) have to do the same activity for a long time (sometimes more than 4 hours without a break). The assembled

Worker 2

Scheme of assembly line.

Human Factors and Ergonomics in Manufacturing & Service Industries

DOI: 10.1002/hfm

157



Figure 3

Quality inspection and indexes in assembly process.

assembly line (ppmself insp ) and then during the inner electrical inspection (ppmel-insp ). The indicator ppminsp does not include failures detected in installation of wire harnesses in cars or those detected during operating the car by the driver (ppmcust ). Yet, the research shows that this indicator is significantly smaller than ppminsp . Research hypotheses: •

•

From many various WF only a few have a significant impact on the QAP. (Although this statement seems to be quite obvious, recognizing these factors in the particular process is not common in practice.) There are considerable interactions between WF with respect to QAP, which can magnify or weaken the direct impact of the individual factors taken separately.

3. DETERMINATION OF THE SET OF THE MOST SIGNIFICANT WF The research was initiated with a survey conducted on approximately 100 assembly workers to determine what, in the employees’ opinion, are the main reasons

Figure 4

158

for failure. The respondents were asked to rank the possible failure reasons on a scale from 1 to 10, with 10 as the most important. For each reason, the mean rating and its range were calculated. The failure reasons are shown in Figure 4 in the sequence corresponding to the employees’ ratings. It can be noticed that, according to employees’ statements, the main reasons for assembly failures are as follows: manual work, employee fatigue, time pressure (hurry), noise level (NL; perceived by the human working on the assembly line), work monotony, faulty elements, employee inattention, microclimate, not respecting work instructions, and lighting. All of these factors (excluding faulty items, employee inattention, and not respecting regulations) belong to the WF. • •

The material environment factors are the following: noise, microclimate, and lighting. Techno-organization factors are the following: handwork (method in Figure 1), hurry, fatigue, and monotony (rhythm, pace, and intervals in Figure 4).

The results obtained from the employee survey were confirmed in an additional analysis performed by a team of experts using Failure Mode and Effect Analysis (FMEA) methodology (Benbow et al., 2002; Kowalik, 2008). Despite the fact that the team members performing FMEA were not aware of the survey outcome, they came to conclusions similar to those obtained from the employees survey (Kowalik, 2008). By choosing WF for further research the following criteria were considered: 1) How can they be controlled? and 2) How can they be changed without any additional investment (the experiment was conducted

Rating of assembly failures according to employees.


DOI: 10.1002/hfm



in practical conditions without influencing the production process)? Taking into consideration both WF criteria mentioned above, only two were used in the research: NL and work monotony. From a practical point of view, both can be controlled relatively easily and effectively. (This is explained in the Results section.) The factors of manual work, employee fatigue, and time pressure ought to be recognized as uncontrollable and associated (implicitly) with work monotony. They are inherent characteristics of the harness assembly process and cannot be reduced with the use of simple organizational measures. Indeed, the following can be assumed: • •

•

Manual work is the essence of wire harness assembly and cannot be eliminated. The fatigue level of the fitters results from individual characteristics and the work monotony itself. Work under high time pressure is a result of a bad production or shop organization.

Lighting and microclimate, although significant and easy to control, were neglected because in workshop reality their changes are not possible without substantial technical investment.

4. RESULTS Both of the WF (work monotony and NL) in real conditions are controlled to a limited level only. Besides, work monotony is difficult to measure and to quantify. Therefore, it was assumed that work monotony is represented by the variation of the production program, referred to as production variation (PV). PV is understood as the number of various types of wires assembled by the fitters working at a separated assembly line during a single shift. It can be observed that the lower the number of types of wires produced within one shift, the higher the work monotony. The PV can be controlled by proper production timetable planning. Like work monotony, the NL perceived by the human working on the assembly line can also be controlled in natural manufacturing conditions only to a certain extent. It can be reduced by the use of special protective screens as well as by proper production organization and elimination of the loudest machine tools at the same time (Engel et al., 2000). An example of noise field at assembly lines being investigated is presented in Figure 5. The individual assembly lines (S1, S2, . . . S7) are exposed to noise

Figure 5

Noise fields at assembly line.

ranging from 66 to 80 dB. It can be noticed that the highest NL exists at work stations situated close to the noisiest machine tools or other noise sources. In the real production conditions presented, designing an optimal experiment is a quite complex task. Therefore, it was decided to design a passive experiment in which the controlled factors are fixed at the levels that are enabled by the production program and not at the levels that are determined by the researcher autonomously (without taking into consideration environment constraints). On the one hand, this kind of experiment can be viewed as being disadvantageous because of the limited researcher’s influence on its design. The researcher must adapt the experiment to the possibilities offered by the production conditions in the shop. On the other hand, the passive experiment makes the results more convincing because they are obtained under natural circumstances. In this kind of experiment the employees do not realize that their performance—with respect to the experiment response variable—are measured and evaluated. The idea of the designed experiments is shown in Figure 6 (Benbow et al., 2002). The controllable factors, which are the subject of the investigation, are work monotony and NL. These factors affect the human working on the assembly line and indirectly affect the QAP, which is quantified by the index ppm, the


DOI: 10.1002/hfm

159



Controlable factors factors under investigation - Work monotony (production variation – PV) - Noise level – NL (at work stations on the assembly line) Input conditions Input material Assembly method People and their skills Production organization

Response variable- „ppm insp” (Human work quality quality of the assembly process QAP)

Assembly process

Uncontrolable factors - time pressure - material faults - others

Figure 6

Variables and factors of the experiment.

•

response variable. The other WF are treated as uncontrollable. The goal of the main experiment was to answer two questions: 1) Is the influence of PV and NL on the QAP significant? and 2) Is the interaction between PV and VL in respect of QAP significant? The controllable factors were set at three levels (with respect to all constraints resulting from the passive nature of the experiment design): •

Three levels of noise: NL1 = 66–70 dB, NL2 = 70–75 dB, or NL3 = 76–80 dB (see Figure 6).

The results of the experiments are presented in Table 1. Due to the fact that the above-mentioned experiments were performed in natural production conditions, the results obtained are disturbed by many uncontrolled factors (see Figure 4). Therefore, the data have been collected in such a way that the analysis of variance (ANOVA; Aczel, 2000) was possible. For each combination of the investigated factors PV and NL, 24 observations of ppminsp were been completed.

Three levels of PV: PV1, PV2, and PV3, respectively with one type, two types, or three types of wire harness produced within one shift; and

TABLE 1. Factorial Experiment with Two Factors: PV and NL. NL Level 1

PV Level

2

3

1

316; 265; 282; 294; 450; 621; 325; 302; 364; 440; 563; 440; 549; 258; 564; 251; 672; 256; 457; 589; 587; 427; 301; 439

564; 437; 657; 517; 1393; 718; 405; 802; 625; 564; 1175; 800; 755; 665; 874; 645; 1354; 1002; 1428; 700; 964; 1212; 600; 924

2

198; 214; 207; 207; 181; 284; 298; 306; 299; 293; 157; 149; 292; 381; 390; 396; 120; 362; 149; 263; 268; 102; 163; 100 170; 95; 99; 98; 112; 152; 169; 266; 184; 146; 190; 267; 109; 118; 164; 254; 187; 254; 124; 473; 103; 164; 123; 76

209; 181; 163; 157; 159; 150; 273; 309; 301; 429; 282; 361; 361; 337; 273; 354; 500; 155; 253; 320; 200; 233; 124; 102 142; 124; 123; 131; 394; 425; 145; 147; 130; 121; 167; 139; 167; 168; 195; 211; 230; 238; 100; 123; 170; 154; 108; 164

3

857; 1258; 985; 1180; 1483; 834; 1356; 855; 796; 1498; 753; 626; 1200; 1460; 1500; 1300; 785; 1036; 1100; 850; 1377; 609; 1268; 1069 730; 599; 378; 466; 402; 539; 727; 722; 614; 415; 684; 407; 418; 591; 405; 710; 722; 540; 655; 264; 321; 340; 645; 245 287; 110; 145; 144; 281; 353; 238; 221; 319; 140; 343; 297; 387; 276; 103; 297; 332; 186; 387; 287; 332; 340; 178; 354

Results are in ppminsp .

160


DOI: 10.1002/hfm


Figure 7


Quality index ppminsp as a function of PV and NL.

The results of the experiment are shown in Figure 7. The influence of PV on ppminsp seems to be more distinct than the influence of NL. Also, it is noticeable that the impact of PV is stronger at the lower NL. To show the significance of the relationships shown in Figure 7, three pairs of hypotheses were put forward: 1)

2)

3)

H0 : Factor PV has no influence on ppminsp versus H1 : Factor PV has an influence on ppminsp ; H0 : Factor NL has no influence on ppminsp versus H1 : Factor NL has an influence on ppminsp ; and H0 : There is no interaction between PV and NL with regard to ppminsp

versus H1 : There is an interaction between PV and NL with regard to ppminsp . All the hypotheses were tested at the level of significance α = 0.05. The results of classical ANOVA calculations are presented in Table 2, where the sum of square, mean square, and degrees of freedom are necessary to calculate the F0 ratio (Aczel, 2000). Because the value of F0 for NL is smaller than F critical (p > 0.05) there are no grounds to reject the zero hypothesis in reference to this factor. In statistical terms, this means that PV has no significant influence on ppminsp . For hypotheses referring to factor PV and interaction between PV and NL, the p value is less than 0.05. It


DOI: 10.1002/hfm

161


Impact of Selected Work Condition Factors TABLE 2. Analysis of Variance Results Source of Variation NL PV Interaction NL × PV Error ∗

Sum of Squares

Degrees of Freedom

Mean Square

60,116.4 176,954.1 1,198,908.1 5,892,878.4

2 2 4 207

52,257.65 26,217.42 276,378.8 16,429.54

Fo

p Value

Critical F for α = 0.05

1.056 3.108∗ 10.529∗

0.349 0.043∗ 0.000000087∗

3.04 3.04 2.42

The significant value of the statistics F and p value.

means (from statistical point of view) that the influence of both of these factors is significant.

5. DISCUSSION As a result of the experiments, the following conclusions can be formulated: •

•

•

Work monotony expressed by the PV has a considerable impact on the quality of the researched assembly type, which means with significant participation of a human. Increasing production variety of the assembly program allows for the reduction of the number of assembly failures. This influence, however, can be useful only to a certain level of PV. After exceeding a certain value of PV, the work quality would decrease. The impact strength of NL depends on the PV. It is significant only from a certain noise intensity, however. Up to this noise threshold value (in this experiment, ca. 70 dB), its influence is negligible. A significant interaction between NL and PV was observed. It means that the increasing assembly process variety is accompanied by the increased workers’ resistance against the noise. This statement can be generalized and applied to other WF, such as manual work, the employee’s fatigue, and time pressure connected with work monotony. This observation is most valuable because it allows for optimizing these WF from the process quality point of view and adopting them to the current work conditions.

6. SUMMARY The ability and readiness of the human involved in manufacturing processes to do the job correctly are 162

strongly influenced by work environment conditions. The results of the investigations described in this article are that noise and work monotony, among other factors associated with work environment, have a particularly important influence on the human and the quality of work. Moreover, these factors show significant interaction affecting human ability and concentration. It should be taken into account while optimizing the environmental conditions of the production process.

References Aczel, A. A. (2000). Complete business statistics, (In Polish). Statystyka w zarz„dzaniu. Wydawnictwo Naukowe PWN, Warszawa. Benbow, D., Berger, R., Elshennaway, A., & Walker, H. (2002). The certified quality engineer. Milwaukee, WI: ASQ Quality Press. Bradley, G., & Hendrick, H. W. (1994). Human factors in organizational design and management. Amsterdam: North-Holland. Broberg, O. (2007). Integrating ergonomics into engineering: Empirical evidence and implications for the ergonomists. Human Factors and Ergonomics in Manufacturing, 17(4), 353–366. Cheng, O. S., & Jiang, B. C. (1995). Integrating safety in manufacturing systems. Human Factors in Manufacturing, 5(3), 303–324. Engel, Z., Piechowicz, J., & Stryczniewicz, L. (2000). Podstawy wibroakustyki przemysłowej (in Polish). Indus´ trial vibroacoustics. Akademia Gorniczo – Hutnicza, ´ Wydział In˙zynierii Mechanicznej Robotyki, Krakow. Fitts, P. M. (1951). Engineering psychology and equipment design. In S. S. Stevens (Ed.), Handbook of experimental psychology. New York: Wiley. Helander, M. (2006). A guide to human factors and ergonomics. London: Taylor & Francis Group. Karwowski, W., & Salvendy, G. (1994). Organization and management of advanced manufacturing (pp. 76–77). Toronto: John Wiley & Sons.


DOI: 10.1002/hfm


Kowalik, D. (2008). Badanie wpływu wybranych czyn´ ´ nikow Srodowiska i organizacji pracy na jako´sc´ procesu produkcyjnego z dominuja¸cym udziałem pracy człowieka (in English: Influence of chosen environmental and organisational factors on quality of production process with significant human participation). Praca doktorska. Poznan´ University of Technology. Mantura, W. (2008). Human factors in quality management. Human Factors and Ergonomics in Manufacturing, 18(5), 565–579. Senge, P. (2008). The necessary revolution. How individuals and organizations are working together to create a sustainable world. NY: Doubleday Publishing Group. Singleton, W. T. (1974). Man-machine systems. New York: Penguin Press. Singleton, W. T., Fox, J. G., & Whitfield, D. (1971). Measurement of man at work: An appraisal of physiological and psychological criteria in man-machine systems. London: Taylor & Francis. ´ Springer, K., & Gorny, A. (2008). Use of ergonomic solutions in industrial safety formation. Foundations of Control and Management Sciences, 11. Stanton, N. A., Hedge, A., Brookhuls, K., Salas, E., & Hendrick, H. W. (2004). Handbook of human factors and ergonomics method. Boca Raton, FL: CRC Press.


Szulczyk, J., & Cempel, C. (2006). Redukcja hałasu na linii produkcjipłyt warstwowych (in English: Noise reduction during production process of composite panel). ´ Zeszyty Naukowe Politechniki Poznanskiej, Wydział ´ Budowy Maszyn i Zarza¸dzania, no 3, Poznan. Tytyk, E. (2004). Evolutional background for humanizing technology. Human Factors and Ergonomics in Manufacturing, 14(3), 307–319. Tytyk, E. (2008). Management of work conditions quality through ergonomic designing of work condition sources. Foundations of Control and Management Sciences, 11. ´ Wasinska, K. (1999). Jako´sc´ s´ rodowiska pracy i jej wpływ na funkcjonowanie człowieka w systemach technicznych (in English: Work environment quality and its influence on human in technical systems). Wy˙zsza ´ Szkoła Pedagogiczna im. Tadeusza Kotarbinskiego. ´ Zielona Gora. Wilson, J. R., Corlett, E. N., & Manenica, I. (1987). New methods in applied ergonomics. London: Taylor & Francis. ˙ ´ Zurek, J., Ciszak, O., Cie´slak, R., & Suszynski, M. (2007). Metodyka projektowania ergonomicznego stanowiska monta˙zu re¸cznego (in English: Methodology of designing the ergonomic-based manual assembly work´ station). Zeszyty Naukowe Politechniki Poznanskiej, ´ Wydział Budowy Maszyn i Zarza¸dzania, no 7, Poznan.


DOI: 10.1002/hfm

163