Crawler Crane Maintenance Optimization with Increased Reliability Through Preventive and Corrective Maintenance Strategies

Crawler cranes are a type of heavy equipment that plays a crucial role in various industrial sectors, especially in construction projects, heavy lifting, and other activities that require high lifting capacity and mobility in difficult work areas. The operational reliability of crawler cranes is a critical factor in supporting the smooth running of business processes, considering that downtime or failure of this equipment can result in project delays, cost losses, and safety risks. Therefore, it is essential to ensure that maintenance of this equipment is carried out effectively and efficiently [1-3]. However, the challenge in crawler crane maintenance is achieving an optimal balance between preventive maintenance (PM) and corrective maintenance (CM) strategies. The PM strategy aims to prevent damage through periodic maintenance, replacement of spare parts, and routine inspections. On the other hand, the CM strategy is carried out when components are damaged or not functioning correctly, which aims to repair or replace problematic components [4, 5].

PM has become a significant focus on improving equipment reliability and operational efficiency in recent decades. Several studies have shown various approaches that can be adopted in the PM strategy, either through applying new technologies, developing innovative methodologies, or integrating with the production process. The research conducted by Hardt et al. [6] proposed the application of a modified Total Productive Maintenance (TPM) methodology by utilizing industry 4.0 technology to improve PM on production equipment. Differently, Wang et al. [7] developed a Product-Service System (PSS) approach for active maintenance of complex equipment, which integrates services and products in one system to improve efficiency. On the other hand, Li et al. [8] used deep learning methods in highway maintenance, showing how automated decision-making can support data-driven PM programs. Then, another algorithm-based approach was proposed by Yang et al. [9], who used reinforcement learning to optimize production and maintenance schedules in multi-component systems. Dui et al. [10] adopted a cost-based approach in industrial robot maintenance, emphasizing the balance between cost and reliability. On the other hand, Gholizadeh et al. [11] investigated PM on a waste-to-energy system by considering uncertainty in production scheduling.

Li et al. [12] conducted other studies focusing on maintenance scheduling and developed a PM schedule optimization based on production-maintenance synchronization. Liu et al. [13] investigated parallel machines' integrated production and maintenance scheduling to reduce production disruptions. Wocker et al. [14] complemented these studies by introducing flexibility in job shop scheduling that considers PM.

In addition, risk- and cost-based approaches have also received attention in the literature. Nasrfard et al. [15] proposed a probabilistic optimization model for maintenance inspection rates, considering the correlation between maintenance cost, duration, and state transition probability. Zhen et al. [16] investigated the maintenance interval setting on safety barriers in offshore installations, considering risk and cost. Hernández-Chover et al. [17] analyzed the comparison of PM costs and improvements in asset management to improve operational efficiency.

Integration between maintenance and inventory management is also a topic explored in depth. Zhang et al. [18] studied the optimization of maintenance and inventory management in a standby system with damaged spare parts. Afifi et al. [19] developed a memetic-based algorithm for simultaneous management of maintenance scheduling and spare part inventory. Furthermore, PM strategy optimization in a multi-component system was carried out by Zhang et al. [20], who proposed a method to overcome the competing risk of interdependent s-components. Alamri and Mo [21] optimized a PM regime based on a failure system model that considers reliability. Hamdan et al. [22] studied maintenance optimization in a k-out-of-n weighted repairable system, which aims to improve reliability. Other studies, such as Al-Refaie and Almowas [23], emphasize the importance of multi-objective maintenance planning in improving operational efficiency. Wang et al. [24] developed an optimal condition-based maintenance policy for a balanced system, emphasizing the importance of balancing performance and cost in maintenance strategies. Kim et al. [25] explored PM optimization in the electricity market, focusing on balancing reliability and cost.

However, most of the studies that have been conducted tend to focus on one aspect of maintenance or combine maintenance with production schedules without integrating PM and CM strategies in depth. In addition, the improvement of reliability and costs for maintenance activities carried out by previous studies has not achieved a balance between reliability and costs that reaches a very good reliability level. Unlike those studies, this study aims to optimize crawler crane maintenance by increasing reliability through the implementation of balanced PM and CM strategies. Optimization is carried out by analyzing existing maintenance data, identifying critical systems that often fail, and determining the ideal PM interval. With this optimization, crawler cranes are expected to operate more reliably and efficiently and reduce the risk of unplanned downtime. In addition, this study will also evaluate the impact of the proposed strategy on maintenance costs, downtime duration, and overall system reliability. Thus, the results of this study are expected to provide practical guidance in crawler crane maintenance management for related industries.

3.1 Initial data collection

The research was conducted at a contractor company in Indonesia engaged in the Engineering Procurement Construction Installation (EPCI) sector. The company uses a crawler crane on the pipelay barge to load and unload materials from the barge to the barge or vice versa. Data from the operation and maintenance of crawler cranes will be collected and processed from January 2022 to August 2024. The following is a table of crawler crane maintenance and operation data.

In Table 1, crawler crane failures were identified in the hydraulic, lifting, and electric systems. Therefore, these systems require special attention. High-cost repairs replace critical components such as hose and wire rope hoist (hoist and boom). PM is carried out on lifting and mechanical system components to reduce the risk of more severe but irregular failures. Working hours based on Indonesian time consist of 24 hours, starting from 00.00 to 24.00. Company activities start from 08.00 to 16.00, but 12.00 to 13.00 is a break time, so it is not counted as operating hours. During operating hours other than break time, even though the crawler crane is on standby, it is declared uptime because it is ready to use. Based on operating hours from 08.00 to 12.00, 4 working hours are obtained and continued from 13.00 to 16.00, 3 working hours are obtained, so the total working time available for the crawler crane per day is 7 working hours. The expenditures due to maintenance are PM and CM costs, including labour costs, spare parts, and other costs. The total combined maintenance cost of PM and CM is Rp. 126,260,000. While the PM cost itself is Rp. 315,000, and the CM cost is Rp. 125,945,000.

Table 1. Crawler crane maintenance and operation data

3.2 Initial data calculation

3.2.1 Calculation of mean time between maintenance (MTBM)

It is known that the total available time is 6832 hours, with operating conditions of 6738 hours and maintenance conditions of 94 hours, with 11 maintenance activities combined between PM and CM. The total maintenance cost due to PM and CM is Rp. 126,260,000. The average cost of each maintenance is obtained from the total maintenance cost divided by the number of maintenance so that the result is Rp. 11,478,182. Therefore, the MTBM for the crawler crane unit is:

MTBM $=\frac{\sum \text { TBM }}{\mathrm{N}_{\text {failure }}}=\frac{6738 \text { hours }}{11}=612,55$ hours

3.2.2 Calculation of mean time to maintenance (MTTM)

The total repair time of the crawler crane is 94 hours. The number of maintenance is 11 times. So, MTTM for crawler crane is:

MTTM $=\frac{\sum \mathrm{T}_{\text {repair }}}{\mathrm{N}_{\text {failures }}}=\frac{94 \text { hours }}{11}=8,55$ hours

3.2.3 calculation of achieved availability (Aa)

The calculation of initial availability follows the equation of achieved availability (Aa). The calculation results are as follows:

$\mathrm{Aa}=\frac{\mathrm{MTBM}}{\mathrm{MTBM}+\mathrm{MTTM}}=\frac{612,55}{612,55+8,55}=98,62 \%$

3.2.4 Initial reliability (R) calculation

The initial reliability calculation follows the general reliability equation, which is exponential. The calculation results are as follows:

$\mathrm{R(t)}=\mathrm{e^{-\lambda t}}=\mathrm{e^{-0,0016 \times 612,55}}=36,79 \%$

3.3 Analysis of operation and maintenance data with a probability distribution

A probability distribution analysis is carried out for crawler crane units to determine the distribution using operation and maintenance data. Furthermore, parameter estimation is carried out to optimize MTBM, MTTM, Aa, R values, and maintenance costs. Parameter analysis and estimation using Minitab 18 software with the following results (refer to Table 2).

1a.jpg

(a) Probability plot for TBM

1b.jpg

(b) Distribution overview plot for TBM

Figure 1. Goodness-of-Fit probability plot TBM and distribution overview for TBM

2a.jpg

(a) Probability plot for TTM

2b.jpg

(b) Distribution overview plot for TTM

Figure 2. Goodness-of-Fit probability plot TTM and distribution overview for TTM

Table 2. Parameters according to probability distribution and calculation of MTBM, MTTM, Aa, and R

Based on the probability distribution test shown in Figure 1 and Figure 2, the results show that the TBM data is normally distributed, and the TTM is Lognormal distributed. Then, the MTBM calculation was carried out with the results of 612.55 hours and MTTM for 8.70 hours. Based on the MTBM and MTTM values, the Aa results were 98.60%, and R was 50%. The results of the MTBM, MTTM, Aa, and R calculations show that only the R-value increased from 36.79% to 50%. While the other values are still the same, so optimization is needed to obtain the optimal value.

3.4 Development of the PM interval optimization model

The PM interval involves maintenance costs. It is known that the PM cost (C_PM) per incident is Rp. 78,750 obtained from Rp. 315,000 / 4 times maintenance. The CM cost (C_CM) per incident is Rp. 17,992,143 obtained from Rp. 125,945,000 / 7 times maintenance. As for calculating the optimal PM interval for the crawler crane unit, if using the MTBM value of 612.55 hours as a guide, the preventive maintenance interval is carried out every 529.05 hours. Therefore, optimization is needed to increase R by using the PM interval formula at different times (t):

$\mathrm{T}_{\mathrm{PM}}=\sqrt{\frac{2 \times \mathrm{C}_{\mathrm{CM}} \times \mathrm{MTBM}}{\mathrm{C}_{\mathrm{PM}}}}=\sqrt{\frac{2 \times 17.992 .143 \times 612,55}{78.750}}=529,05$ hours

3.5 Simulation and optimization iteration

The simulation uses different times of MTBM values. The simulation stage is carried out to validate the results of the PM interval calculation with the total cost and R system. Simulations are carried out with various variations of PM intervals to determine the minimum total cost. The formula for total maintenance costs (C_total):

$\mathrm{C}_{\text {total }}=\mathrm{C}_{\mathrm{PM}}+\left(1-\mathrm{R}\left(\mathrm{T}_{\mathrm{PM}}\right)\right) \times \mathrm{C}_{\mathrm{CM}}$

For example, at TPM at 529.05 hours, the reliability result R(T_PM) is 50.00%. Then, the total cost is

$\mathrm{C}_{\text {total }}= 78.750 + \left( {1 - 50\%} \right) \times 17.992.143 = \mathrm{Rp.~9.074.821}$

The calculated value of C_total at TPM 529.05 hours is still too large. Therefore, iteration needs to be done by changing the T_PM value to find the minimum total cost balanced with the R-value. This iteration starts from time (t) 119 hours and its multiples.

3.6 Model testing and discussion of data processing results

Furthermore, iteration refers to the provisions, namely time (t) 119 hours and its multiples. Here are some choices of PM intervals considered in the iteration:

Table 3. Iteration choices as considerations for T_PM, R, and C_total

3.jpg

Figure 3. Comparison of TPM, R, and C_total values (t = 6832)

From the choices in Table 3 and Figure 3, the 320-hour interval is chosen as the optimal interval because it provides a balance between high R and low C_total and balanced frequency maintenance. Furthermore, model testing is carried out by comparing the results before and after optimization, calculating the MTBF, MTTR, reliability, and total maintenance costs. The comparison results before and after optimization are shown in Figure 4.

4.jpg

Figure 4. Comparison of C_total, R, MTTM, and MTBM before and after optimization

Figure 4 shows that maintenance optimization successfully increased R from 36.79% to 90.04 while reducing Ctotal from Rp. 126,260,000 to Rp. 39,947,255. The cost reduction is the result of simulation and iteration carried out by conducting several maintenance trial scenarios at certain times, as shown in Table 3 and Figure 3. Furthermore, the MTTM value increased to 8.7 hours, while MTBM decreased from 612.55 hours to 320 hours. This reduction in MTBM is to increase T_PM with the aim of more routine and scheduled PM so that it can prevent unexpected CM actions. Based on the results of data processing, this study is better than the traditional PM and CM planning methods that only rely on irregular PM schedules and CM carried out only when there is damage. In this study, the application of PM interval optimization can increase the R of the crawler crane and reduce the total maintenance cost. The optimal PM strategy is generated through historical data analysis, optimization model development, and simulation and iteration that consider R and costs. This research is something new that does not only focus on one aspect of maintenance or combines maintenance with production schedules but is more in-depth by integrating PM and CM strategies in depth along with consideration of R and total costs, which are calculated by conducting simulations with the same total time or duration as historical crawler crane maintenance data, namely 6832 hours.