Elsevier

Safety Science

Volume 62, February 2014, Pages 205-213
Safety Science

Obstacle clearance while performing manual material handling tasks in construction sites

Highlights

•

We analyse how manual material handling task may contribute to fall accidents in construction tasks.

•

Changes in obstacle clearance were examined according to load weight and handling position adopted by the worker.

•

Load weight affects the obstacle clearance and increases the possibility of falling.

•

Load handling strategy adopted by the worker does not affect the obstacle clearance.

•

Manual material handling tasks imposes, per se, significant changes in the process of obstacle clearance.

Abstract

Construction is widely known as having high rates of fall accidents. In spite of constant technological advances and increasing process automation, manual material handling still takes place in many construction tasks. These two factors yielded the following research question: “May manual material handling contribute to fall accidents?”. The aim of the research reported in this article is to evaluate the likelihood of same-level falls while performing tasks involving manual material handling during obstacle clearance in various handling positions.

A laboratory-based study was performed through the simulation of manual material handling tasks using a 4 m long treadmill. Eight construction workers participated in this study. Participants were tested in three different load positions holding different load weights (10 kg, 18 kg, and 25 kg) while walking on the treadmill.

The results demonstrated that the obstacle clearance pattern changes due to the load weight, however, no influence was observed on the load handling strategy. This variation of pattern increases the probability of tripping and falling. Recommendations were made in order to prevent falls in construction sites while performing manual material handling tasks.

Keywords

Health and safety
Manual material handling
Obstacle clearance
Construction
Accidents

1. Introduction

According to literature, the construction industry is the sector with the highest number of work-related accidents, falls from height being the most common phenomenon (Ale et al., 2008, Chi, 2004, Chi et al., 2005, Goldsheider et al., 2002, Hämäläinen et al., 2006, Haslam et al., 2005, Hinze et al., 1998, Hinze et al., 2005, Huang and Hinze, 2003, HSE, 2009, Janicak, 1998, Jeong, 1998, Koningsveld and Molen, 1997, Pan et al., 2003; Perttula et al., 2003, Salminen, 1995). However, the importance of accidents due to same-level falls should not be underestimated. Although they are less frequent, they may cause significant injuries to construction site workers (Glazner et al., 2005, Lipscomb et al., 2006). Falls on the same level are associated with trips resulting from the typical lack of organisation in construction sites (Glazner et al., 2005, Lipscomb et al., 2006).

In spite of constant technological advances and process automation, there are several construction industry tasks which involve frequent manipulation of various items (Pan et al., 2003). These items differ in their characteristics (Paquet et al., 1999), and they can reach considerable size and weight, which make their handling difficult (Lipscomb et al., 2006, Pan et al., 2003, Smallwood, 2006), disrupting the worker’s postural balance (Kollmitzer et al., 2002, Pan et al., 2003).

Ambient vision is primarily responsible for the perception of motion and spatial orientation and thus is also important for control of balance (Tang and Woollacott, 2004, Cinelli and Patla, 2008) and perception of risk factors. The size of the load and the way it is being loaded in construction sites, may obstruct workers’ views, which may increase the chances of tripping situations and lead to the occurrence of falls, namely during obstacle clearance. In fact, Mohagheghi et al. (2004) concluded that visual information plays an important role during the obstacle clearance process. According to these authors estimating the spatial location of the obstacle during the approach phase is critical to adjust foot placement during obstacle clearance process (Mohagheghi et al., 2004).

Manual material handling simultaneously with working in the same position for long periods; bending or twisting the back in an awkward way; working in awkward or cramped positions and working when injured or hurt, is presented as one of the five most important safety, health and ergonomic problems in construction industry (Smallwood, 2006). Operations involving manual material handling of loads are very common in construction, specially in tasks involving scaffolding, formwork, structural steelwork, masonry, building fabric, plumbing, suspending ceiling and paving (Albers and Estill, 2007, Smallwood, 2006, Molen et al., 2005, Hess et al., 2003, HSE, 2000, HSE, 2003, Goldsheider et al., 2002, Elders and Burdorf, 2001, Beek et al., 2000, Paquet et al., 1999, Heran-Le Roy et al., 1999).

A research study conducted by Lipscomb et al. (2006) showed that 11.5% of work accidents involving tripping and slipping were related to manual handling of loads and in 24.4% of these accidents the dimension and weight of the load had a significant contribution to the occurrence of falls.

According to scientific literature there are, at least, three options for obstacle clearance, one can step over, step on, or circumnavigate the obstacle (Austin et al., 1999, Berard and Vallis, 2006). Stepping over an obstacle seems to place the greatest demand on the locomotor system and poses the greatest risk of falling, (Austin et al., 1999). This option is also frequently used by workers on the construction site based on the results of direct observation.

The risk of tripping or falling, during obstacle clearance, occurs due to the potential clash over the obstacle with the toe or heel of the lead limb or the unstable and protracted stance phase in which the centre of mass is outside of the narrow base of support (Austin et al., 1999).

There are a plethora of studies involving the analysis of obstacle clearance process (Eng et al., 1994, Patla and Prentice, 1995, Taga, 1997, Chou and Draganich, 1998, Austin et al., 1999, Gonçalves et al., 2000; Pijnappels et al., 2001; Hess and Dietz, 2003; Mohagheghi et al., 2004, Berard and Vallis, 2006, Fink et al., 2007). However, the contribution of manual material handling to the occurrence of falls during obstacle clearance especially in construction industry has been little dealt with, and thus has become a research opportunity.

This paper presents an exploratory study through laboratory simulation of some manual material handling tasks on the construction site with obstacle clearance, with the participation of eight construction industry workers, and it aims to understand the contribution of these tasks to accident occurrence.

2. Methods

2.1. Participants

Eight volunteer construction workers, with ages comprised between 25 and 45 years old, participated in this study. Their mean age was 36 (±7.52) years, mean weight 78.5 (±6.16) kg and mean height 1.73 (±0.04) m. Subjects were informed on the research procedures before they gave informed consent. All the participants were selected from construction companies and from the same professional category.

The participants were required to undergo a health-history screening before joining the study to ensure that the participants did not have any of the following conditions: history of dizziness, tremor, vestibular disorders, neurological disorders, cardio pulmonary disorders, diabetes, chronic back pain, chronic knee pain, chronic joint pain and any fall resulting in an injury requiring days off work within the past year. A medical history containing one or more of these conditions could influence the participant’s performance during testing.

2.2. Protocol

Reflective body markers using double-sided adhesives were placed bilaterally on the skin, overlying five anatomical points: hallux, calcaneus, lateral malleolus, femoral condyle and head of the greater trochanter (Fig. 1).

  1. Download : Download full-size image

Fig. 1. Illustrative scheme of the anatomic points were the marks were placed.

During the tests, the workers walked continuously on a 4-m long treadmill (MIRALAGO make, built specially for the purpose) at a constant speed of 1.1 m/s (4 km/h), corresponding to normal gait speed. The gait was considered comfortable by all participants at this speed. Several trials were performed in construction sites comprising of normal gait on the floor with obstacle clearance. During these trials the subjects were instructed to walk a known distance at a self-selected velocity time spent by each subject was registered and then the velocity was calculated. The results obtained by these trials confirmed the adequacy of the treadmill speed for the performance of trials.

Three obstacles were placed on the treadmill belt, at variable time intervals and distances. The obstacles were hidden by a curtain to avoid obstacle anticipation during the gait process; as a consequence, subjects did not know whether an obstacle was positioned. Prior to the tests, all participants performed walking exercises in order to get used to walking on the treadmill.

The obstacles’ heights were chosen on the basis of their relevance to daily activities in construction sites, as a result of direct observation of construction workers performing their tasks in loco. All obstacles had 220 mm height, 500 mm width and 310 mm depth. Reflective markers were attached to the four corners of both sides of the obstacle.

The workers were instructed to step over the obstacles. To better simulate working conditions, all participants wore a safety helmet and safety boots. The participants wore a full-body safety harness which was attached to a ceiling mounted hook to insure that subjects would not become injured should their recovery reaction be inadequate.

The experimental procedure comprised of three trials, with a maximum duration of 1 min and a half, in four different tasks performed by the following order:

•

Walking on the treadmill with obstacle clearance without carrying a load.

•

Walking on the treadmill with obstacle clearance carrying a 25 kg load.

•

Walking on the treadmill with obstacle clearance carrying a 10 kg load.

•

Walking on the treadmill with obstacle clearance carrying a long sized 18 kg load.

A standard 45 cm (length) × 30 cm (width) × 10 cm (height) box was used to hold the 10 kg and 25 kg loads, which were evenly distributed in the box.

During manual material handling tasks, the workers adopted three distinct postures, which are quite common in the construction industry:

•

Shoulder loading.

•

Lateral loading.

•

Frontal loading.

The 18 kg load was not frontally handled due to its excessive length (150 cm length × 15 cm width × 10 cm height).

The loading strategies were selected based on the information of eleven focus groups involving a total of 44 participants covering a range of six different stakeholders from the industry, such as Construction Site Supervisors, Ergonomists, Civil Engineers, Safety Engineers, Occupational Medicine Doctors and Health and Safety Coordinators. During the interview, each group was asked to consider the main strategies adopted by workers while performing manual material handling tasks in construction sites. This information was complemented with observation of manual material handling tasks in construction sites.

The 10 kg and 25 kg loads were equivalent to carrying cement bags and the 18 kg was equivalent to carrying floor planks, once the length of this load is of higher dimension than the other two loads.

2.3. Data collection and analysis

The subjects were videotaped performing three trials for each of the four tasks. Gait kinematics were collected through VICON Motion Capture Software, at 240 Hz, for further biomechanical analysis of the following variables (Fig. 2), which are typical of the obstacle clearance process (Austin et al., 1999, Berard and Vallis, 2006, Mohagheghi et al., 2004, Patla, 2004):

–

“Toe clearance” – Minimum vertical distance between the marker of the toe and the upper proximal corner of the obstacle (TC).

–

“Heel clearance” – Minimum vertical distance between the marker of the heel and the upper distal corner of the obstacle (HC).

–

Horizontal distance between the foot and the obstacle before clearing the obstacle (HDBO), at toe-off.

–

Horizontal distance between the foot and the obstacle after clearing the obstacle (HDAO), at heel contact.

  1. Download : Download full-size image

Fig. 2. Illustrative scheme of the parameters analysed during the tests HDBO – horizontal distance from the foot to the obstacle before transition; HDAO – horizontal distance from the foot to the obstacle after transition; TC – “Toe clearance”, minimum vertical distance between toe and obstacle; HC – “Heel clearance”, minimum distance between heel and obstacle.

Except for “heel clearance”, which was only collected for the lead foot, the remaining variables were collected for both feet: the lead foot (foot that leads the obstacle clearance process) and the trail foot (foot that sustains the obstacle transition process).

One way analysis of variance (ANOVA) on all factors was performed for all dependent variables. Comparisons were made between load weight, load handling strategy and obstacle clearance variables. Pairwise post hoc analysis (Tukey’s H.S.D. test) of the ANOVA results was performed to determine which means were significantly different from each other. Significant level was set at p = 0.05.

3. Results

Table 1 shows the mean and standard deviation results for all eight workers, which was obtained for the following variables for both feet, during obstacle clearance:

•

“Horizontal distance from the foot to the obstacle before transition” (HDBO).

•

“Horizontal distance from the foot to obstacle after transition” (HDAO).

•

“Toe clearance” (TC).

•

“Heel clearance” (HC).

Table 1. Mean values obtained for Horizontal distance from the foot to obstacle before transition (HDBO); Horizontal distance from the foot to obstacle after transition (HDAO); “Toe clearance” (TC) and “Heel clearance” (HC), for the lead limb and for the trail limb.

TrialHDBO (mm)HDAO (mm)TC (mm)HC (mm)
MeanSdMeanSdMeanSdMeanSd
Lead foot
No load792.5212.2129.560.5229.832.4183.121.6
Shoulder loading (10 kg)637.6155.7155.955.8159.537.5145.238.7
Lateral loading (10 kg)610.0126.1147.441.3146.023.6172.149.2
Frontal loading (10 kg)695.1200.2156.963.7176.815.2160.140.8
Shoulder loading (25 kg)562.1174.5137.459.3145.820.5146.318.6
Lateral loading (25 kg)634.4148.4135.739.3154.127.1143.125.8
Frontal loading (25 kg)599.0100.8184.982.7161.419.1163.824.7
Shoulder loading (18 kg)609.6202.1155.840.2183.121.7156.345.6
Lateral loading (18 kg)582.9106.5165.543.6139.123.8152.439.3
TrialHDBO (mm)HDAO (mm)TC (mm)
MeanSdMeanSdMeanSd
Trail foot
No load241.8102.0663.1149.7250.865.2
Shoulder loading (10 kg)142.383.1800.178.1121.471.3
Lateral loading (10 kg)115.868.0763.3106.3148.897.2
Frontal loading (10 kg)133.060.5814.480.7150.478.9
Shoulder loading (25 kg)110.525.6709.970.179.9648.0
Lateral loading (25 kg)125.144.2700.165.0107.966.9
Frontal loading (25 kg)137.983.2712.9140.7107.143.6
Shoulder loading (18 kg)112.267.6774.737.982.153.9
Lateral loading (18 kg)82.923.7780.587.873.135.2

3.1. Influence of load handling strategy on obstacle clearance

ANOVA results revealed no statistical differences for any dependent variable in all load handling strategies adopted by the workers (p > 0.05), for either foot (Table 2).

Table 2. ANOVA results and Tukey’s HSD test for the various load handling strategies, depending on the load weight carried.

Shoulder loadingLateral loadingFrontal loading
ANOVAH.S.D. TukeyANOVAH.S.D. TukeyANOVAH.S.D. Tukey
Lead foot
HDBO0.151Inexistence of groups0.005⁎2 Groups: (0); (10, 18, 25)0.010⁎2 Groups (0); (25)
HDAO0.056Inexistence of groups0.124Inexistence of groups0.091Inexistence of groups
TC0.000⁎2 Groups: (0); (10, 18, 25)0.000⁎2 Groups: (0); (10, 18, 25)0.000⁎2 Groups: (0); (10, 25)
HC0.013⁎⁎2 Groups: (0, 18); (10, 18, 25)0.035⁎⁎2 Groups: (0, 10, 18); (10, 18, 25)0.12Inexistence of groups

Trail foot
HDBO0.000⁎2 Groups: (0); (10, 18, 25)0.000⁎2 Groups: (0); (10, 18, 25)0.000⁎2 Groups: (0); (10, 25)
HDAO0.000⁎3 Groups: (0, 25); (25, 18); (18, 10)0.001⁎2 Groups: (0, 25); (10, 18, 25)0.001⁎2 Groups: (0, 25); (10)
TC0.000⁎3 Groups: (0); (10, 25); (18, 25)0.000⁎2 Groups: (0); (10, 18, 25)0.000⁎2 Groups: (0); (10, 25)
⁎

P < 0.01.

⁎⁎

P < 0.05.

3.2. Influence of the load weight on obstacle clearance

3.2.1. Lead foot

ANOVA results revealed significant statistical differences for HDBO when carrying lateral loads and frontal loads. In both cases two homogeneous groups were found as shown in Table 2.

Fig. 3(a) corresponds to the performance of HDBO as a function of load weight for the various handling strategies adopted by the worker.

  1. Download : Download full-size image

Fig. 3. (a) HDBO behaviour according to load weight and the adopted handling strategy; (b) behaviour of TC (lead foot) depending on the load weight and load handling strategy adopted; and (c) behaviour of Heel Clearance (supporting foot) depending on the load weight and load handling Strategy.

The analysis of Fig. 3(a) reveals that HDBO seems to decrease as the load increases. Results indicate a greater proximity of the foot to the obstacle before transition with an increase in load for all the strategies adopted for load handling. Exceptions are the mean values obtained for 18-kg during frontal load and 25-kg during shoulder load.

However, chart analysis and post hoc analysis uncover significant differences among the mean values obtained for the pairs 0 kg and 10 kg, 18 kg, 25 kg during lateral load and for the pairs 0 kg and 25 kg during frontal load.

The mean value of HDBO obtained for an 18-kg load during lateral load is lower than for a 25-kg load. This result is inconsistent with initial expectations, as a lower value was expected for the heaviest load (25 kg). Thus, as the main difference between the 18-kg load and the 25-kg load is related to its length, this variation seems to influence the workers’ behaviour when stepping over the obstacle.

In view of the results obtained for HDBO, it can be stated that the increasing of load weight showed a tendency to place the foot closer to the obstacle, an effect that is statistically significant (p < 0.05). This is the case for a frontal load of a 25-kg load and for lateral load of 10-kg, 18-kg and 25-kg loads when compared to the performance of the task without any load.

As for the behaviour of HDAO, there are no statistically significant differences in any of the strategies adopted for load handling (P > 0.05). Based on these results for the lead foot, it can be stated that variable HDAO was not influenced by load weight. As a result, the worker’s behaviour was not affected while stepping over the obstacle.

The analysis of Table 2 also reveals statistically significant differences in the value of TC during lateral, frontal or shoulder loading (P < 0.01). The results obtained through Tukey’s H.S.D. test indicate the existence of two homogeneous groups for shoulder, lateral and frontal loading, that is, (0 kg) and (10 kg, 18 kg, and 25 kg) for shoulder and lateral loading and (0 kg) and (10 kg and 25 kg) for frontal loading.

Fig. 3(b) describes the behaviour of the Toe Clearance variable, for the lead foot, according to load weight, for the various load handling strategies adopted by the worker.

A significant decrease is found in the Toe Clearance value when obstacle clearance includes manual material handling. Although there is not a statistically significant difference in the Toe Clearance values obtained for the various loads, a tendency to a decrease in the Toe Clearance value is found when load weight increases. A more obvious effect was found during frontal and shoulder loading. As a result, it can be stated that carrying a load while stepping over an obstacle brings about a decrease in Toe Clearance.

Regarding the behaviour of Heel Clearance, the analysis of Fig. 3(c) as well as ANOVA analysis, show the existence of statistically significant differences in the values obtained for these variables during shoulder loading and lateral loading, (P < 0.05). The results from H.S.D. Tukey’s test indicate the existence of two homogeneous groups for shoulder loading (0 kg and 18 kg) and (10 kg, 18 kg, and 25 kg) and two homogeneous groups for lateral loading (0 kg, 10 kg, and 18 kg) and (10 kg, 18 kg, and 25 kg) the inclusion of the same load in both groups can be explained by the variance of the values obtained in certain loads.

Fig. 3(c) presents a chart which describes the behaviour of the Heel Clearance variable for the lead foot as a function of load weight, for the various handling strategies adopted by the worker.

The change in Heel Clearance with load weight is not uniform. Some fluctuation can be seen especially during frontal loading.

A statistically significant decrease in HC is found only when carrying 25 kg during lateral loading, when compared with the performance of the same task without a load.

Shoulder loading of 10 kg and 25 kg loads also presents significantly lower values of HC as compared to the ones obtained without load handling.

As a consequence, it can be stated that manual material handling causes a decrease in Heel Clearance, exception should be made for the values obtained for an 18 kg load that showed higher Heel Clearance.

3.2.2. Trail foot

ANOVA analysis showed a significant variation in HDBO between loads during shoulder lateral and frontal loading (p < 0.01). In these cases, there are two homogeneous groups as Table 2 illustrates.

The chart presented in Fig. 4(a) replicates the results from ANOVA. HSD Tukey’s test shows that there is a decrease in HDBO with load handling. The behaviour of this variable for the trail foot is quite similar to that of the lead foot.

  1. Download : Download full-size image

Fig. 4. (a) Behaviour of HDBO (trail foot) depending on the load weight and the load handling strategy adopted and (b) of HDAO (trail foot) depending on load weight and handling strategy adopted; Behaviour of TC (trail foot) depending on the load weight and the handling strategy adopted.

Thus, it can be seen that load handling produces closer foot proximity to the obstacle, which does not depend on load weight. This effect is clear in all of the strategies used for load handling.

Although it is not statistically significant, according to these tests, there is a tendency towards a decrease in the value of HDBO as the load increases for shoulder and lateral loading. However, contrary to initial expectations, which anticipated a lower value for a 25 kg load, the HDBO value was lower for an 18 kg in both strategies. As the 18 kg object was larger than the others, it is likely that the object length influences negatively the worker’s performance in obstacle clearance.

During frontal loading, load handling produces a decrease in the value of HDBO when compared with obstacle clearance without a load.

Based on ANOVA results, HDAO changes for the trail foot are significant (p < 0.01) during shoulder loading, lateral loading and frontal loading. In these cases there are three homogeneous groups for shoulder loading and two homogeneous groups for lateral and frontal loading, as Table 2 demonstrates.

Fig. 4(b) presents the variation of HDAO for the trail foot as a function of load weight, for the various load handling strategies.

We can see an increase in HDAO for a 10 kg load. This phenomenon is statistically significant for lateral or frontal loading. For higher values, there is a decrease in HDAO, that is, closer foot proximity to the obstacle. This decrease is statistically significant with a 25 kg load during shoulder loading.

Analysis of Fig. 4(a) and (b) shows a reciprocal behaviour of HDBO and HDAO for the trail foot. When the value of HDBO increases, the value of HDAO decreases and vice versa. This behaviour is consistent with the biomechanics of walking, as it deals with the same foot while performing a step.

As to the behaviour of TC, the ANOVA analysis showed that the variation between loads is significant during shoulder loading, lateral loading and frontal loading (p < 0.01). In these cases there are three homogeneous groups for shoulder loading and two homogeneous groups for lateral loading and frontal loading, as shown in Table 2.

Fig. 4(c) describes the behaviour of Toe Clearance for the trail foot, depending on load weight, for the different handling strategies adopted by the worker.

The value of Toe Clearance decreases when obstacle transition includes load handling. This phenomenon is statistically significant for any of the postures adopted by the worker.

Although there is not a statistically significant difference in the Toe Clearance values obtained for the various loads, there is a decreasing tendency depending on the increase of load weight.

The 18 kg load, regardless of the handling strategy adopted by the worker, obtained the lowest value of Toe Clearance, this fact being statistically significant during shoulder loading. This behaviour contradicts the initial expectations which anticipated a lower value for the 25 kg load. Bearing in mind that the major difference between the two loads relates to the fact that the 18 kg load was considerably larger than the 25 kg load, we can state that the length of the object exerts an additional effect in disturbing the obstacle clearance process by construction workers.

The results obtained for Toe Clearance for the trail foot are quite similar, in terms of variation, to the results obtained for the same variable for the lead foot. This situation is due to the fact that the preceding transition by the lead foot acts as a reference to positioning the trail foot (Mohagheghi et al., 2004, Patla, 2004).

4. Discussion

In this study individuals had to step over obstacles while carrying a load. To perform this task successfully Toe Clearance (TC), Heel Clearance (HC), Horizontal distance from the foot to the obstacle before transition (HDBO); horizontal foot to the obstacle after transition (HDAO) are the key measures to avoid accidental contact with the obstacle.

In this experiment the weight load carried by the individuals and the load carrying strategy was systematically changed.

The discussion of results is organised around the influence of the load weight on obstacle clearance.

4.1. Influence of the load handling strategy on obstacle clearance

The differences obtained for the variables do not seem to depend significantly on the load handling strategy adopted by the worker (handling posture). These results are not in accordance with the initial expectations since it was expected that frontal loading would have imposed major difficulties in the transition process due to the forward shift of the overall system (body and load) with a migration of the Centre of Gravity to the front, generating more forward and downward forces, as the body begins to roll over the heel. These results can be explained by the limitations of load weight (Hsiang and Chang, 2002), since there is an ergonomic limitation for the load weight carried by the individuals, although in construction sites workers usually handle heavier loads than those tested in this study.

On the other hand these findings may be explained by the fact that in frontal loading the load was carried next to the body trunk. As a consequence the frontal roll effect of the body did not reveal too much expression.

4.2. Influence of the load weight on obstacle clearance

The results showed that load handling produces a variation in the obstacle transition pattern involving a decrease in the horizontal distance before obstacle clearance for both feet, an increase of the horizontal distance after thee obstacle for the trail foot and a decrease of the vertical distance (Toe Clearance and Heel Clearance) for both feet.

Despite not noticing a variation in the previous variables it should be stated that some of these variables are not affected by the increase of the load weight (e.g. HDBO and HDAO). However, these variables are affected when the task is performed carrying a load or without carrying loads. Once again, these results can be explained by the limitations of the load weight, imposed by ergonomic principles.

According to the results, the individuals had the need to change gait path to better place the foot before obstacle clearance. This strategy is in accordance with the results reported by Eng et al., 1994, Taga, 1997 and Austin et al. (1999) and is more pronounced with the increase of load weight. This is understandable since it increases postural stability during the transition of the obstacle.

Chou and Draganich (1998) and Mohagheghi et al. (2004) reported that decreasing the horizontal distance before obstacle clearance influences the decrease of the Toe Clearance during obstacle transition. When we compare the changes in these variables during the trials we can find a similar phenomenon, which is related to the increase of load weight.

The similar behaviour obtained for Toe Clearance in both feet is in accordance with Patla (2004) since the lead foot acts through proprioceptive stimulus as a guide to the position of the trail foot.

Furthermore the results obtained for 18 kg load, specially the ones achieved for HDBO for both feet and TC for trail foot contradicts the initial expectations, which foresaw a lower value with a 25 kg load. This fact may be linked to the length of the object.

4.3. Limitations

A few limitations ought to be considered as a result of the implemented methodology. Although they were consciously accepted, they affect the final results, thus requiring special attention in the interpretation and generalisation of the results.

Among these limitations one stands out: the fact that the experiment used a treadmill – which offers more stability to the walking process. However, the lack of space in the lab to better carry out the experiment, the limited scope of the digital image cameras, the need for reproducibility of the tests, and the impracticality of performing the trials in a construction site did not allow for implementing pre-defined paths. Using a treadmill offers the advantage of being able to analyse the obstacle transition strategy in a more detailed way, as well as measuring subsequent steps which are essential to regaining balance after tripping over the obstacle. There are several differences between stumbling on the treadmill as opposed to the floor (Cordero et al., 2003). In the first the subject is forced to keep to the speed of the treadmill when dealing with the perturbation. The swing leg is perturbed but still moves forward with the treadmill band that transports the stance limb. This is different from stumbling on the floor (Cordero et al., 2003). As reported by Cordero et al. (2003)) recovery responses found during the trials agreed with those described in the literature for hitting obstacles, either on the ground or on the treadmill, (Eng et al., 1994, Schillings et al., 1996, 2000; Troy and Grabiner, 2005, Cordero et al., 2004).

The reduced size of the sample also affects the statistical tests. In spite of this, the study found that the subjects who participated in the tests carried out the obstacle transition in a similar way. The age of the participants may also be a limitation since age seems to increase the probability of tripping and reduce the ability of balance recovery after a trip (van Dieën et al., 2005). However, scientific literature points out that these limitations are quite evident when subjects have more than 60 years old (Tang and Woollacott, 2004, Overstall, 2004, Kenny and Armstrong, 2004, Van Dieën et al., 2005, Kim et al., 2005, Butler et al., 2006). Additionally, the results of the focus groups previously performed all interviewees agreed that manual material handling tasks are performed by younger workers.

The present study took place in a laboratory which does not recreate, in full, the environment of the building site. However, this seems to be one of the problems affecting most of the studies which involve ergonomic analysis in the construction industry. The ephemeral nature of the tasks as well as the continuous alteration in the workplace makes it impossible to carry out ergonomic studies in situ. These are some of the reasons for the scarcity of ergonomic studies in this sector.

5. Conclusion

We can conclude that manual material handling contributes to the occurrence of falls during obstacle clearance. As the load weight increases, there is a tendency for a closer proximity of the foot to the obstacle, either horizontally or vertically. This effect is more obvious with heavier loads, increasing the risk of tripping and also the possibility of a fall.

The results of this study emphasise the need of intervention measures in order to prevent falls in construction sites, namely:

–

Manual handling operations should be avoided where reasonable to do so, and those which do take place should be limited as far as reasonable. Therefore, a reduction in the amount and extent of manual handling should be sought for every task, specially in what concerns with the load weight, which should be limited.

–

Improve the housekeeping and organisation of the construction site to minimise risks. This can be done through the maintenance of a tidy site and the removal of obstacles that may contribute to tripping hazards, as well as providing stable walkways.

–

Lighten the load by using packages/containers which are smaller, easier to manage and easier to grasp.

–

Prevail the use of precast elements since it will reduce not only the number of workers in contraction sites, as well as manual material handling tasks.

This study also highlights the need to plan manual material handling tasks in the pre-construction phase during the overall project risk assessment in order to minimise the need for manual handling of loads, the reduction of distances where manual handling is required and the removal of obstacles whenever this tasks must be performed.

Once the study seems to indicate that the occurrence of falls is also influenced by the length of the object and considering that the great majority of studies in manual material handling, in construction industry, give more importance to the weight of the load than its length, which maximum lifting values are well studied it is suggested further investigation on the effect of load’s length in the maintenance of postural balance, specially during obstacle clearance.

References

1

Tel.: +351 229866081.

2

Tel.: +351 253 510 170/1.

View Abstract
Copyright © 2013 Elsevier Ltd. All rights reserved.