Optimization of Wiring Harness Logistics Flow in the Automotive Industry

Abstract

This paper presents a compelling argument for optimizing the logistics flow of wiring harnesses within the automotive industry to address the rising production demands of vehicle manufacturers. It introduces an innovative assembly line structure specifically designed to boost efficiency and enhance responsiveness to client needs. Drawing from data gathered from an actual assembly line dedicated to producing engine harnesses for K9K engines, this study offers a practical and impactful foundation for its proposed optimization strategies. The new assembly structure effectively merges the benefits of a dynamic line—which emphasizes efficient space utilization and flexibility—with the strengths of a rotary line, particularly in light of the increasing complexity associated with harness production. The paper features a mathematical model that calculates cycle times for workstations within this new system architecture, optimizing the entire production process. Moreover, it illustrates how advanced modeling, simulation, and optimization techniques using WITNESS Horizon Version Release 25.0 can identify necessary adjustments for achieving optimal assembly line balance. Additionally, this research addresses pressing environmental concerns by proposing a robust recycling strategy for the scrap produced during wiring harness manufacturing. By advocating for sustainable practices and responsible waste management, the study highlights the importance of minimizing the ecological footprint of the automotive manufacturing process. In summary, this research provides essential insights and practical solutions for optimizing wiring harness logistics flow in the automotive industry. By implementing these strategies, manufacturers can significantly enhance their production capacity, improve operational efficiency, and maintain competitiveness in an ever-evolving market landscape.

1. Introduction

In the contemporary landscape of automotive manufacturing, the efficient management of supply chain logistics stands as a critical determinant of operational success. Within this context, optimizing wiring harness logistic flow emerges as a critical point for enhancing production efficiency and cost-effectiveness.

The optimization of logistics flow for wiring harnesses presents a multifaceted challenge that demands a comprehensive approach. This article delves into the complexities of wiring harness logistics within the automotive industry, aiming to provide a comprehensive understanding of the challenges involved in proposing innovative strategies for optimization and ultimately improving the overall competitiveness of automotive manufacturers. This study offers industry professionals and researchers practical insights and innovative solutions by synthesizing engineering, logistics, and management principles. Through a comprehensive analysis, this work seeks to contribute to the ongoing efforts to refine the operational dynamics of automotive supply chains, ultimately fostering enhanced competitiveness and sustainable practices within the industry.

Previous research was centered on analyzing the main aspects related to the harness production process, assembly flow optimization [1], challenges in increasing productivity by integrating automation solutions and proposing a model of RFID implementation in wiring component storage systems [2].

The current research addresses challenges associated with augmenting production capacity, illustrated through a case study involving a wiring harness producer tasked with doubling its production capacity. Furthermore, it will explore sensitive topics such as recycling waste materials.

The wiring harness industry plays a crucial role in the automotive sector, providing essential components that enable the efficient and reliable transmission of electrical signals and power within complex systems. Wiring harnesses, also known as cable harnesses or wiring looms, consist of insulated wires, connectors, terminals, and protective sleeves bundled together to form a unified and organized assembly. These harnesses are custom-designed to suit specific applications, and are crucial for ensuring the seamless operation of electrical and electronic systems. As such, they are fundamental to the functionality and performance of a wide range of vehicles, aircraft, and machinery.

The wiring harness industry encloses diverse activities, including the design, engineering, manufacturing, and testing of harnesses tailored to meet the unique requirements of various industries. This involves selecting appropriate materials, such as insulated wires, connectors, and protective sheathing, as well as the precise arrangement and bundling of these components to ensure optimal performance, reliability, and durability in demanding environments. Additionally, the industry also involves compliance with stringent regulatory standards and quality control measures to ensure safety, reliability, and conformance to industry-specific requirements.

Advancements in technology and manufacturing processes have significantly influenced the wiring harness industry, developing specialized harnesses that incorporate advanced materials, miniaturized components, the monitoring of process parameters, and increased functionality [3,4]. The vehicle assembly line’s design affects the wiring harness’s definition because some subassemblies (e.g., doors) must be separately wired before they are finally mounted on the car [5].

Furthermore, the industry has seen a shift towards environmentally sustainable practices, including using recyclable materials and implementing eco-friendly manufacturing processes to reduce the environmental impact of harness production.

The wiring harness industry is characterized by a high degree of standardization and regulation, as the safety and reliability of wiring harnesses are paramount in ensuring the overall integrity of the systems in which they are installed. As a result, manufacturers must adhere to stringent quality standards and regulatory requirements, often necessitating rigorous testing and certification processes. Furthermore, the wiring harness industry is subject to continuous technological innovation and advancements. This includes developing new materials, such as high-performance insulating and shielding materials, as well as integrating advanced technologies, such as data transmission capabilities and electromagnetic interference (EMI) shielding.

Additionally, the industry is increasingly focused on sustainability and environmental responsibility, driving the development of eco-friendly materials and manufacturing processes. However, comparing the process assembly flow from twenty years ago in relation to the contemporary process of assembly flow implemented in wiring harness manufacturing, we can observe an important similarity between then and now; even though the assembly tables and electrical check table have evolved, the lines are mainly following the same structure, with a dynamic assembly line or rotary line (carousel), depending on the complexity of the wiring harness, and manual assembly operations still represent over 90% of the wiring harness price.

Optimizing the processes is necessary to increase the activity’s added value in this context. Math elements are widely used to describe different technological processes. The Simplex algorithm can be effectively used to control the values of the technological process parameters [6].

Engineering optimization is a global subject of interest for many scientific research teams. It is part of today’s mathematical modeling and approaches to the control of processes and systems. An effective approach is applying nonlinear programming methods (NLP) to optimize solving a specific problem in engineering practice [7].

Also, the Simplex algorithm, a linear programming element by nature, can optimize a process to be profitable [8].

The use of WITNESS Horizon simulation software to optimize production systems within the enterprise can effectively simulate processes and workflows [9].

WITNESS Horizon simulation can dynamically adjust parameter changes in real time. These changes impact the simulation results and improve their accuracy [10].

Integrating the simulations made with the WITNESS Horizon software with the Simplex algorithm represents a powerful approach for analyzing and optimizing complex systems. This integration obtains valuable information that improves the company’s operational performance.

In the context of this paper, the Simplex algorithm is used as a complementary analysis tool to support the results obtained with the help of the WITNESS Horizon software.

Previous research has highlighted the impact of having a production process mainly composed of manual assembly operations on wiring harness assembly productivity, as well as the challenges the harness makers meet in increasing the automation degree. Wenjing and Shichang also studied the impact of having a production process composed mostly by manual operations, and after a deep analysis of each operation, they applied the lean production concept and IE method to identify the bottlenecks and optimize a wiring harness assembly line. They applied the ECRS principles to reorganize each workstation, reduce the unnecessary waiting time and apply the needed layout modifications. The improved production process generated a production rate increase of 22.18%, and required a reduction of six operators [11].

In analyzing research papers related to wire harness assembly and human–robot collaboration (HRC), several key themes and correlations emerge that highlight advancements in manufacturing efficiency and workplace ergonomics, as follows:

• Human–robot collaboration—Several studies emphasize the integration of robots into manual assembly processes, with a focus on enhancing productivity and safety. Heisler et al. presented optimization strategies for wire harness assembly through HRC, showcasing the benefits of combining human dexterity with robotic precision [12]. Similarly, Palomba et al. explored redesigning manual workstations to foster collaborative environments, indicating that a mechatronic approach can facilitate smoother interactions between humans and robots [13];
• Task allocation and workflow optimization—Salunkhe et al. delved deeper into the specifics of task allocation within automotive wire harness assembly stations, addressing how clearly defined roles can optimize human–robot interactions [14]. This aligns well with the findings of Heisler et al., who also proposed methods to improve task efficiency through collaborative practices [12]. Together, these studies highlight that the effective allocation of tasks can significantly enhance assembly line performance;
• Technology integration—The application of augmented reality (AR) in manufacturing, explored by Szajna et al., provides an additional layer of support in the assembly process. While not directly focused on robotics, the use of AR can enhance the capabilities of human workers by providing real-time guidance and improving cognitive processing [15]. This technological integration complements HRC by equipping workers with tools that optimize their performance alongside robotic partners, thereby contributing to a more efficient workflow, as outlined in the other papers;
• Efficiency and ergonomics—Across these studies, there is a consensus on improving the ergonomic aspects of wire harness assembly. By redesigning workstations and task allocations, both human comfort and operational efficiency can be enhanced. The collaborative nature of the systems discussed allows for a more sustainable production approach, reducing the physical strain on workers while maximizing output.

In summary, these papers collectively illustrate the importance of HRC in modern manufacturing, particularly in wire harness assembly. They advocate for integrated technologies, optimal task allocation, and ergonomic workstation designs that contribute to a more efficient and effective manufacturing process.

Increasing the productivity of wire harness production by integrating new technologies is very challenging, and a part of this challenge is given by the wiring harness’s complexity, which incorporates many different components assembled to serve as the nervous system that facilitates the transmission of electrical power and signals to various components and systems. To respond to this challenge, Bi et al. applied a modularization concept to automate the testing processes of cable harnesses for an SME; modularization is aimed at decoupling the complexity of mechanical, electrical, electronic, and control components [16]. Besides power distribution and signal transmission, it connects components throughout the vehicle, making coordinating and synchronizing various functions possible. The wiring harness is designed to function regardless of the harsh environmental conditions within the vehicle (proximity to sharp edges, temperature variations, moisture, vibration, and exposure to chemicals). Research conducted in an aircraft industrial context proposed an A*-ACO algorithm, developed based on the A* algorithm for minimal path detection and the ant colony optimization algorithm, to improve wiring harness layout. Tests done on industrial examples showed an improvement in adapting to the fabrication constraints and reducing the loops [17].

Other studies focus on integrating AI techniques to improve manufacturing outcomes in wiring harness production. Nguyen et al. explored the enabling of deep learning through synthetic data, specifically in the context of automotive wiring harness manufacturing. This approach allows for the generation of diverse training datasets, which is crucial for enhancing the performance of AI models [18]. Similarly, Song et al. presented a fault detection system leveraging AI, demonstrating how machine learning algorithms can be employed to identify defects in the wiring harnesses effectively [19]. Nguyen et al. highlight the role of synthetic data in overcoming challenges related to limited real-world datasets, which can improve the robustness of deep learning models [18]. On the other hand, Song et al. underlined the application of AI for real-time fault detection, which relies heavily on the quality and variety of data available [19]. Together, these studies suggest that both synthetic data generation and effective data utilization are pivotal for achieving quality improvements in manufacturing. The application of these AI-driven methodologies in the automotive sector, as depicted in both studies, underscores the growing trend towards automation and intelligent manufacturing. However, there are some limitations related to data availability, regarding synthetic or real-world datasets, that can cover the multitude of all real possible scenarios emerging in a complex manufacturing environment.

Another challenge in the path of automation is that wiring harnesses are tailored to the specific layout and requirements of different vehicle models, accommodating variations in features and options. To summarize, without going into details, the difficulty in increasing the automation degree is related to the following aspects: difficulty in handling because the wires that are part of the harness are very flexible elements and have long lengths; the large variety of components; the diversity of operations necessary to ensure connectivity; the diversity of the harnesses produced within a project (a large number of references that require different assembly boards); changes to the wiring architecture that occur during the development phase of the project, which involves the rapid modification/adaptation of the order of operations and the assembly system; changes imposed by after-sale quality incidents, although smaller than in the development phase, also involve the rapid adaptation of the wiring assembly system [1].

Our Previous Research

The paper [1] overviews the production process, from design to production. In Figure 1, a schema of the logistics flow of a dynamic assembly line is presented to give a better view of the variety of processes carried out in harness production.

The diagram highlights the logistics flow of the dynamic assembly line, providing an overall picture of how manual and automatic processes intertwine in producing wiring harnesses. The processes carried out in the “Picking zone” and on the assembly line are primarily manual. A mix of manual and semiautomated operations is performed in the preassembled elements production area, the ultrasonic welding area, and the test center. The only part of the process that is fully automated in current practice is represented by the fully automated crimping and twisting machines.

Considering the share of manual operations within the production process, the article presented a theoretical case study to determine the main factors influencing the productivity of dynamic assembly lines in wiring harness production. The simulation carried out in the WITNESS Horizon Software highlighted the importance of equilibrating the tasks assigned to each workstation while considering the operators’ dexterity level. Workstations subjected to substantial workloads necessitate the attendance of experienced operators (with high dexterity). These workstations should be strategically positioned throughout the assembly line to optimize workflow fluidity and diminish waiting periods [1].

Another context that could benefit from a prior simulation and analysis in WITNESS Horizon Software is when the harness-maker must increase the production capacity to respond to their client’s demand. However, expanding the physical space for production may only sometimes be feasible due to cost, logistical, or infrastructure constraints. Thus, the challenge is to find viable methods for augmenting the production capacity of wiring harnesses within the existing spatial confines while ensuring operational efficiency and cost-effectiveness.

One approach to enhancing production capacity within the same space involves optimizing manufacturing processes. As was shown in the previous case study, this implies streamlining workflow, reducing cycle times, and eliminating non-value-added activities. Even though process optimization can support capacity increases, there are some limitations related to the ratio between the old client’s demand and the new ones. If the ratio is around 0.5 or higher, capacity increase only by process optimization is unlikely to be achieved. In this case, modifying the assembly line structure and adding new workstations might be the only viable solution.

The efficient utilization of available space through layout redesign can also contribute to increased production capacity. This involves reconfiguring the production floor to minimize material movement, reduce blockages, and enhance workflow efficiency. By employing advanced layout planning techniques, such as cellular manufacturing and optimized material flow, manufacturers can effectively maximize the utilization of available space while accommodating higher production volumes.

In a preceding article [2], we formulated a theoretical framework outlining a storage system designed to address the distinctive requirements of component warehousing for harness manufacturers in the automotive industry. A Radio Frequency Identification (RFID) Ultra-High-Frequency (UHF) system was structured for deployment across all warehouse sections. The objective was to provide a straightforward and adaptable RFID system variant, characterized by low investment costs, suitable for integration into virtually any preexisting warehouse.

Another challenge that must be addressed is recycling the scrap from wire harness manufacturing, which is a vital process in vehicle production. However, this manufacturing process inevitably generates scrap material, including metals such as copper, aluminum, steel, plastic, and other materials [20]. Recycling this scrap is essential to minimize waste, conserve resources, and reduce environmental impact. This paper will also highlight the importance of recycling wire harness manufacturing scrap and its implications for sustainable production. Recycling scrap material from wire harness manufacturing significantly reduces the environmental impact associated with waste disposal.

Metals such as copper and aluminum are valuable resources that require substantial energy for extraction and processing. Recycling these materials minimizes energy consumption and greenhouse gas emissions associated with mining and refining [21]. The recycling of wire harness manufacturing scrap contributes to the conservation of valuable resources. Copper, in particular, is a critical component in wire harnesses due to its excellent electrical conductivity. Recycling copper scrap reduces the demand for virgin copper, leading to the conservation of natural resources and decreased reliance on mining activities [21,22]. Similarly, recycling plastic components from wire harness manufacturing reduces the consumption of fossil fuels required for plastic production, thereby promoting resource efficiency and sustainability [23,24].

The effective recycling of wire harness manufacturing scrap presents economic advantages. Recycled materials can be reintroduced into the manufacturing process, reducing the need for virgin materials and lowering production costs [25]. Furthermore, the recycling industry generates employment opportunities and contributes to economic growth. Additionally, as the demand for recycled materials increases, there is potential for developing new markets and industries centered around sustainable resource utilization.

While the benefits of recycling wire harness manufacturing scrap are obvious, challenges such as material separation, contamination, and logistical considerations must be addressed. Implementing efficient sorting and processing techniques and collaborating with specialized recycling facilities can enhance the effectiveness of recycling initiatives. Additionally, raising awareness and training employees on the importance of proper waste segregation can optimize recycling efforts within manufacturing facilities.

Prior research mainly focused on improving different aspects of the production process. This study offers a new perspective on wiring harness optimization, aiming more at optimizing existing production processes by taking on the challenge of a new approach and redefining the structure of the assembly line.

The paper is organized as follows: Section 2 includes a description of the proposed methodology and the initial structure of the assembly line’s simulation and analysis. In Section 3, the results are presented and analyzed, while Section 4 discusses the conclusion of the research and future work.

2. Materials and Methods

2.1. Description of the Proposed Methodology

This research aims to optimize the logistics flow of wiring in the automotive industry to meet growing production requirements. This process was optimized through WITNESS Horizon Software. To confirm the efficiency and accuracy of the simulations made through the software, the Simplex Algorithm, a powerful and versatile tool for optimizing industrial processes, was used.

The proposed methodology for this paper, presented below, aims to set goals, identify limitations, and analyses the process before and after optimization.

The analysis is carried out for an assembly line in the automotive industry. The advantages of a dynamic assembly line are also presented, focusing on the efficient use of space and flexibility.

The proposed steps in this methodology are as follows:

• Stage 1—Setting the objectives and limitations accepted in the proposed model;
• Stage 2—Performing the WITNESS Horizon Software for the unoptimized version;
• Stage 3—Diagnosing and performing the simulations in the optimized version;
• Stage 4—Validation of simulations using the Simplex algorithm;
• Stage 5—Analysis of the obtained data.

The objectives accepted in this model refer to optimizing the logistic flow of wiring in the automotive industry.

One issue for improvement in the analysis is that it is focused only on optimizing the logistic flow of wiring without considering the other processes within the organization.

The present model is based on the linear programming problem, through which the extreme point of a real function represented by a linear expression is analyzed when the variables of this function must comply with a set of constraints. Maximum efficiency means minimizing effort and maximizing results, and the concept of optimal is defined, in this case, as $x\in R^n$. This minimises or maximises an objective function. At the same time, it satisfies all the technical and economic constraints.

From the perspective of the existence of a primal-dual pair of linear programming problems, a mathematical model can be developed to simulate industrial phenomena. The Simplex algorithm can only be applied to the standard form of a linear programming problem. However, any canonical form can be transformed into the standard form, and the Duality Theory in the presented model can be adapted to canonical forms.

Solving the two problems in the primal-dual pair can lead to only one of the following three outcomes, according to the fundamental theorem of duality:

a.

If one of the two problems has a finite optimal solution, then the other also has a finite optimal solution, and the values of the objective functions corresponding to the two solutions are equal;

b.

If one of the two problems has an infinite optimum, then the other has no feasible solutions.

c.

If one of the two problems has no feasible solutions, then the other either has an infinite optimum or no feasible solutions.

To justify the statements made earlier, we assume the two standard problems brought to canonical form.

The primary problem is

$$\left\{\begin{array}{c}\left(\max \right)f\left(x\right)\\ Ax\le b\\ x\ge 0\end{array}\right.⟹\left\{\begin{array}{c}\left(\max \right)c^{T}x\\ Ax\le b\\ x\ge 0\end{array}\right.$$

(1)

The dual problem is

$$\begin{array}{c}\left(\min \right)f\left(x\right)\\ A^{T}u\ge c\\ u\ge 0\end{array}⟹\left\{\begin{array}{c}\left(\min \right)b^{T}u\\ A^{T}u\ge c\\ u\ge 0\end{array}\right.$$

(2)

where $c^T$ represents the coefficients of the objective function, $b^T$ represents the constant terms of the constraints, $u^T$ is the vector of the dual problem variables, and A is the matrix of coefficients of the linear combinations among the different constraints.

We will study whether feasible solutions exist for each problem to verify the statements above. Then, for any x feasible solution of the primal and any unfeasible solution of the dual, we have $c^{T}x\le b^{T}u$.

Thus, we have $A^{T}u\ge c$, which is equivalent to ${(A^{T}u)}^T\ge c^T$, and with $u^{T}A\ge c^T$

$$\left\{\begin{array}{c}{u}^{T}A\ge c^T\\ x\ge 0\end{array}⟹u^T\right.Ax\ge c^{T}x$$

(3)

Likewise

$$\left\{\begin{array}{c}Ax\le b\\ u\ge 0⟺u^T\ge 0\end{array}⟹u^T\right.Ax\le u^{T}b \mathrm{s}$$

(4)

From the two systems of Equations (2) and (3), the result is that $b^{T}u\ge c^{T}x$.

In this situation, there are three hypotheses:

• The problem has a finite optimum. In this situation, $x_B$ exists, a basic solution of the primal in standard form, which provides the problem’s optimum, meaning $c^T{x}_B=max c^{T}x$. All minors corresponding to this basis will be positive, resulting in $c_B^T{B}^{-1}A\ge c^T⟺A^T{\left(c_B^T{B}^{-1}\right)}^T\ge c$. This relation indicates that the vector with m components ${u^{\ast }=\left(c_B^T{B}^{-1}\right)}^T$ is an admissible basic solution to the dual problem.

Thus, we have $c^T{x}_B\le u^{T}b$ for any feasible solution to the dual problem. On the other hand, $c^T{x}_B=c^T{B}^{-1}b={\left(u^{\ast }\right)}^{T}b$, from which it results that for any feasible solution to the dual problem, we must check that ${\left(u^{\ast }\right)}^{T}b\le u^{T}b.$ This implies that $u^{\ast }$ is the optimal solution to the dual problem, and that $f\left(x_B\right)=g\left(u^{\ast }\right)$;

2.

The problem has an infinite optimum. In this case, if the dual had admissible solutions u, g(u) would be an upper bound for the set {f(x)/x admissible}, which contradicts the hypothesis of an infinite optimum;

3.

The problem has no solutions. In this case, if the dual had a finite optimum, according to the statements in variant 1, the primal would also have a finite optimum, which contradicts the hypothesis.

Since the three variants cover all possible situations, all leading to one of the previously stated possible outcomes, and considering that it did not matter which problem was chosen for resolution, the previous results have been demonstrated.

According to the complementary slackness theorem, there exists $x^{\ast }{=(x}_{1,}^{\ast }{x}_2^{\ast },\dots x_n^{\ast })$ and $u^{\ast }{=(u}_{1,}^{\ast }{u}_2^{\ast },\dots u_n^{\ast })$, the two admissible solutions of the primal and dual problems, which are optimal solutions for both problems and satisfy the system

$$\left\{\begin{array}{c}{u}_1^{\ast }\left(b_j-{\sum }_{j=1}^n{a}_{ij}{x}_j^{\ast }\right)=0, i=\overline{1,m}\\ \left({\sum }_{i=1}^m{(a}_{ij}{u}_i^{\ast }-c_j\right)x_j^{\ast }=0,j=\overline{1,n}\end{array}\right.$$

(5)

or the matrix-based system

$$\left\{\begin{array}{c}{\left(u^{\ast }\right)}^T\left(b-A{x}^{\ast }\right)=0\\ \left[{\left(u^{\ast }\right)}^{T}A-c\right]x^{\ast }=0\end{array}\right.$$

(6)

It is assumed that ${ x}^{\ast }$ and ${ u}^{\ast }$ are the optimal solutions to the two problems. Therefore,

$$\left.\begin{array}{c}A{x}^{\ast }\le b\\ u^{\ast }\ge 0\end{array}\right\}⟹{\left(u^{\ast }\right)}^T\left(b-A{x}^{\ast }\right)\ge 0$$

$$\left.\begin{array}{c}{A}^T{u}^{\ast }\le c\\ x^{\ast }\ge 0\end{array}\right\}⟹{(\left(u^{\ast }\right)}^{T}A-c)(x^{\ast })\ge 0$$

(7)

Given that ${\left(u^{\ast }\right)}^T=c{x}^{\ast }$(according to the fundamental theorem), ${\left(u^{\ast }\right)}^T\left(b-A{x}^{\ast }\right)+\left({\left(u^{\ast }\right)}^{T}A-c\right)\left(x^{\ast }\right)$ results in ${\left(u^{\ast }\right)}^T\left(b-A{x}^{\ast }\right)=\left({c-\left(u^{\ast }\right)}^{T}A\right)\left(x^{\ast }\right)=0$.

If ${\left(u^{\ast }\right)}^T\left(b-A{x}^{\ast }\right)=\left(c-{\left(u^{\ast }\right)}^{T}A\right)\left(x^{\ast }\right)=0⟹{\left(u^{\ast }\right)}^{T}b=c{x}^{\ast }$, according to the fundamental theorem, $x^{\ast }$ and $u^{\ast }$ are the optimal solutions to the two problems.

The previously described model can be applied within the scope of this study because the time associated with each workstation process depends on previous values. Thus, this simulation aims to optimize time. Applying the previously described model involves establishing three working hypotheses.

2.2. Working Hypotheses

In the analysis process, the times associated with each activity were considered. Thus, we calculated the average for the three processes to identify an equal time value. The three initial postures were as follows: for workstation 1, the time was calculated as 5.5/3 = 1.83333. Since this was period 3, we assigned 1.9 to the first process and 1.8 to the remaining processes. This difference was agreed upon because the first time should be longer than the time for the subsequent two processes. We followed a similar approach for the other workstations.

The constraints used in the first system are the final values corresponding to the first three workplaces. Subsequently, the values for the other workplaces will be established by applying the Simplex algorithm. Thus, the optimal value identified for the first three workplaces will be the final time value for workplace four. The algorithm will continue under the same principle.

The average of the initial times for each process at each workstation was chosen to maximize the function. Thus, for the first iteration, the optimization function consisted of the initial values divided by three (the number of processes). Therefore, for the first workstation, 5.5/3 = 1.8. This was the same for the others.

2.3. Case Study: Increasing Production Capacity by Proposing a New System Architecture

For context, due to customer demand, a harness-maker must double its production capacity within the existing spatial confines. The supplier must also prepare a system that allows efficient recycling of the scrap generated from the production process.

2.3.1. Initial Assembly Line Structure

The initial configuration of the assembly line structure is based on an assembly line used to produce the full version of the engine harness for K9K engines [26] to provide data that can be used in physical implementation.

The line presented in Figure 2 is composed of the following:

• Six fixed workstations are used for wire insertion into the connectors. These workstations are strategically positioned near the initial three assembly boards to facilitate the precise assembly of wires and connectors at a specific stage within the overall assembly process;
• Twenty workstations organized in a rotary assembly line are used for wire layout on the board, making the final subassemblies on the board and adding protections, fixation elements, linked parts, and taping;
• Five workstations are used for clip tests, quality checks, electrical controls, visio tests, and packaging.

In total, 33 operators serve the line:

• There are 6 operators working on the fixed workstations;
• There are 20 operators working on the rotary assembly line;
• There are 5 operators working apart the assembly line in the section dedicated to quality checking, testing (clip, electrical and visio) and packaging;
• There are 2 operators in charge of supplying the workstations with components and preassembled elements.

According to the client’s demand, the initial volume estimation was 38,400 pieces per year. The wiring harness-maker organized the production according to their primary work schedule, which means 48 weeks per year, two shifts per week, and five working days per week.

Summarizing the information above gives us the output production shown in

Table 1. Output production is based on volume estimation and a basic work schedule.

Output Production
800	160	80
Per week (5 days)	Per day (2 shifts)	Per shift

.

One shift consists of 8 h, with a 30 min break. The working time (SWT) within one shift equals 450 min. This information is mandatory for the “Takt time” calculation.

The “Takt time” is a term of German origin (“Taktzeit”) associated with lean management and used in the industrial production environment to refer to the ideal duration of production of a good when it exactly corresponds to the customer’s demand. The production pace must align with “Takt time” to produce precisely the number of units requested by the customer. So, after the production process is decided, the repartition of tasks for every workstation must be made considering the “Takt time” [27].

The “Takt time” is calculated via the following formula:

$$Tt={\displaystyle \frac{SWT}{OP}}={\displaystyle \frac{450}{80}}=5.63 \mathrm{m}\mathrm{i}\mathrm{n},$$

(8)

As shown in Figure 2, the working area is divided into three sections. Each section has its own challenges when adapting to Takt time; this is why task repartition must also be observed in practice, and re-equilibrated if it is the case. The fixed workstation area is prone to the piling up of stocks on one part of the workstation and not being able to provide the necessary items to the rotary line for the workstations that handle more complex operations. For this reason, the complexity of each workstation in this area must be similar to the skill level of the operators. The operators from the rotary section of the assembly line must be well-trained to handle the complexity of the operations while advancing at the pace of the line. They are also the ones that must reach for the wires and other components that are placed behind them. The section where the tests and packaging are apart from the assembly line but must also be synchronized with the “Takt time” of the assembly line to avoid the creation of harness stocks at the workstations.

2.3.2. Initial Structure of the Assembly Line Simulation and Analysis in WITNESS Horizon Software

To obtain data regarding the efficiency of the assembly line’s initial structure and identify aspects that may need adjustments, a simulation of two-shift production was carried out in the WITNESS Horizon Software. WITNESS Horizon was used for modelling the initial system architecture, and a simulation of two shifts of production was carried out to diagnose the key performance indicators associated with the initial assembly line and identify necessary changes to optimize the system as required (to double the production).

The line structure described in Figure 3 is composed of the following:

• FW1–FW6—6 fixed workstations with 6 operators (OP1–OP6);
• R1–R3—preassembled elements trolley;
• Rotary assembly line (conveyor)—20 workstations with 20 operators (OP7–OP26);
• R4–R6—harness trolley;
• Scrap—trolley for rework/scrap;
• CT_QC—clip test/quality check (2 workstations with 2 operators OP27 and OP28);
• EC_VT—electrical control/vision camera check (2 workstations with 2 operators OP29 and OP30);
• Packaging—1 workstation with 1 operator (Op31).

The parameters for each workstation and operator were defined. A diagnosis for the initial system’s configuration was given to identify bottlenecks and long waiting times. The results are presented in Figure 4, Figure 5 and Figure 6.

The results for FW1–FW6 show that six fixed workstations with six operators are blocked for between 7.33% (FW3, FW4, FW6) and 9% (FW1, FW2, FW5) of the total simulation time. For FW1, idle time is 8.33%, and busy time is 82.67%. The busy time for FW2 and FW5 is 91% of the total simulation time, and the busy time for FW3, FW4, and FW6 is 92.67%.

Operators OP7–OP16 are allocated to 10 workstations from the left side of the rotary assembly line. The workstations for OP7 and OP8 are blocked for 8.83% and busy for 91.17% of the total simulation time. The workstations for OP9 and OP10 are blocked for 1.67% and busy for 98.33% of the total simulation time. The workstations for OP11 and OP12 are busy 100% of the time. The workstation for OP13 is idle for 14% and busy for 86%, the workstation for OP14 is idle for 12.33% and busy for 87.67%, the workstation for OP15 is idle for 10.67% and busy for 89.33%, and the workstation for OP16 is idle for 9% and busy for 91% of the total simulation time.

Operators OP17–OP126 are allocated to 10 workstations from the right side of the rotary assembly line. The workstation for OP17 is idle for 10.65% and busy for 89.35%, the workstation for OP18 is idle for 10.45% and busy for 89.55%, the workstation for OP19 is idle for 7.19% and busy for 92.81%, the workstation for OP20 is idle for 8.82% and busy for 91.18%, the workstation for OP21 is idle for 5.56% and busy for 94.44%, the workstation for OP22 is idle for 7.33% and busy for 92.67%, the workstation for OP23 is idle for 9% and busy for 91%, the workstation for OP24 is idle for 7.17% and busy for 92.83%, the workstation for OP25 is idle for 6.17% and busy for 93.83%, and the workstation for OP26 is idle for 5.17% and busy for 94.83% from the total simulation time.

The bottlenecks are caused by the differences between the cycle times defined for each workstation and operator. The results show that a productivity of 132 pieces was obtained for two shifts. The productivity of 132 pieces resulting from the simulation is lower than the estimated 160 pieces. Also, the “Takt time” given by the simulation is 6.80 min, which is lower than the calculated “Takt time” of 5.63 min.

The calculated “Takt time” of 5.63 min represents an average of all busy times for all workstations, while the “Takt time” resulting from the simulation is determined by the main bottleneck and idle time of the workstations, indicated by the workstation with the lowest busy time out of the entire assembly line.

3. Results and Discussion

3.1. Introducing a New System Architecture

Based on the initial system’s diagnosis produced by the flow simulation, the changes needed for the system’s redesign are described and analyzed to remove the bottlenecks, increase the system production capacity, and decrease the “Takt time”. The proposals concerning the characteristics and configuration of each work point are described to obtain an optimized system version.

Considering the need to double capacity, the proposed system architecture aims to combine the advantages of two efficient modes of organizing the assembly line: dynamic and rotary lines. Dynamic assembly lines offer more efficient space usage with reduced and well-defined workstations and frontal alimentation with components. Unproductive times are reduced by having all the components at reach and all the operations made on fixed boards. Integrating a dynamic assembly line makes the system more flexible and reactive towards the client’s needs. If simple modifications are required, the only board that needs to be adjusted is the one that covers the step of the process in which the modification can be applied. Rotary lines offer a better display of the wires, being more efficient in handling the complexity of the wiring harness.

Figure 7 presents a hybrid line structure intended to reap the benefits of the dynamic assembly line in terms of efficient space utilization and flexibility to change and combine with the advantages of a rotary assembly line once the complexity of the harness increases.

The proposed structure consists of two identical assembly lines, as presented in Figure 7.

Each assembly line is composed of:

• Eight fixed workstations, organized into a dynamic assembly line;
• Ten workstations organized in a rotary assembly line are used to handle the wiring harness’s assembly when the structure created in the dynamic assembly line section becomes challenging. In this section, the harness structure is completed with wires, protections, fixation elements, linked parts and taping (most taping operations occurred in this assembly line section);
• Five workstations for clip tests, quality checks, electrical control, vision tests and packaging.

In total, 25 operators serve each line:

• 8 operators working on the dynamic assembly line section;
• 10 operators working on the rotary assembly line;
• 5 operators working apart from the assembly line in the section dedicated to quality checks, testing (clip, electrical and visio) and packaging;
• 2 operators in charge of supplying the workstations with components and preassembled elements.

The production was organized on two identical hybrid lines to double the capacity and optimize space utilization, as shown in Figure 7. As a consequence of the structural change, the tasks have been reorganized. Four pre-blocking workstations have been moved in the prefabricated elements area to produce preassembled elements with longer wire lengths. Two pre-blocking workstations have been placed at the beginning of the dynamic assembly line to handle wire insertion operations to provide the necessary elements for the next workstations. Workstations 3 to 4 are in charge of adding wires to the structure of the harness, insertion operations, forming branches of the harness, and minor tapping operations. The LAD segment of the assembly line features a conveyor belt situated at the base of the workstations. Operators perform snap-in operations by positioning the wires into the clamps affixed to the conveyor.

The new volume estimation communicated by the vehicle manufacturer is 76,800 pieces per year. The work schedule will be the same—two 8 h shifts per day, five working days per week, 48 weeks per year. The estimation by week, by shift and by day is shown in

Table 2. Output production is based on estimating new volumes and the essential work schedule.

Output Production
1600	320	160
Per week (5 days)	Per day (2 shifts)	Per shift

.

The “Takt time” of the hybrid assembly line is 5.63 min (for each of the two identical sides of the proposed assembly line structure).

There are two concepts applied in managing the allocated time for each workstation:

• Open-station concept—In scenarios where the operator is unable to fulfil the allocated task within the specified timeframe, the alternative of relocating the product to the subsequent station is available, effectively mitigating potential backlog. Conversely, should the operator successfully conclude the task before the cycle time elapses, an opportunity arises to work ahead of the schedule. If the total cycle time is exceeded, an alert sound is emitted, and the line is stopped. This concept is intended for rotary assembly lines;
• The closed-station concept is implemented within dynamic assembly lines, where the operator must stop the line if the allocated time is not respected, even in the case of minor delays, as required, for instance, during a straightforward rework procedure [28] (pp. 2–3).

The suggested hybrid structure for reconfiguring the assembly line integrates the two concepts. The assembly line initiates with closed stations within the dynamic segment and transitions to open stations in the rotary segment. This design aims to enhance the management of wiring harness complexity while minimizing the required stops in the event of operational delays.

3.2. Proposed Structure of the Assembly Line Simulation and Analysis in WITNESS Horizon Software

Comparing the results from the initial system simulation with the ones obtained after applying the mathematical model, it can be observed that the mathematical model shows that other optimization solutions might still be available for increasing productivity. Cycle times for the workstations included in the newly proposed system architecture were calculated considering the proposed mathematical model. The solution to converge to a time within the range [5.32; 5.43] minutes is to modify the architecture of the wiring harness. For example, assuming that the harness has a thicker branch (main branch) ending in three or four smaller branches (secondary branches), one branch can be separated from the main branch. In this way, a better equilibration between the workstations can be achieved. However, optimizing assembly time by modifying the wiring harness architecture faces many constraints related to the vehicle environment, like difficulties in ensuring the right path of the harness due to the lack of fixation points in the vehicle or proximity to sharp edges. This is why cycle times for some workstations will not be defined within the range [5.32; 5.43] minutes in the following simulations for the proposed structure of the assembly line.

Two production shifts were simulated in WITNESS Horizon Software to obtain data regarding the proposed structure’s efficiency and to identify aspects that may need adjustments. WITNESS Horizon simulation was used to model the proposed system architecture and diagnose the key performance indicators associated with the proposed system configuration. Using WITNESS Horizon, the system’s configuration was modified to correspond to the new characteristics and to obtain an optimized system version.

The proposed structure of the assembly described in Figure 8 is composed of two lines that follow the same structure, as described below:

• Dynamic assembly line 1—8 workstations with 8 operators (OP1–OP8)/Dynamic assembly line 2—8 workstations with 8 operators (OP26–OP33);
• R1—preassembled elements trolley after dynamic assembly line LAD Section HAL 1/R5—preassembled elements trolley after dynamic assembly line LAD Section HAL 2;
• Rotary assembly line 1 (conveyor)—10 workstations with 10 operators (OP9–OP18)/Rotary assembly line 2 (conveyor)—10 workstations with 10 operators (OP34–OP43);
• R2–R4—harness trolleys after rotary assembly line LAD Section HAL 1/R6–R8—harness trolleys after rotary assembly line LAD Section HAL 2;
• Scrap—trolley for rework/scrap after electrical control hybrid assembly line 1/Scrap01—trolley for rework/scrap after electrical control hybrid assembly line 2;
• CT_QC—clip test/quality check hybrid assembly line 1 (2 workstations with two operators OP19 and OP20)/CT_QC01—clip test/quality check hybrid assembly line 2 (2 workstations with two operators OP44 and OP45);
• EC_VT—electrical control/vision camera check hybrid assembly line 1 (2 workstations with two operators OP21 and OP22)/EC_VT01—electrical control/vision camera check hybrid assembly line 2 (2 workstations with two operators OP46 and OP47);
• Packaging hybrid assembly line 1—one workstation with one operator (OP23)/Packaging hybrid assembly line 2—one workstation with one operator (OP48).

After modeling the assembly line’s proposed structure, the trajectories of the mobile entities were established, and various system parameterization versions were tested based on the values calculated considering the proposed mathematical model; system simulations were performed, and the best system version was identified. Several simulations were performed to optimize the system by changing and testing the parameters for work points and operators. The version with the most significant productivity increase was kept without bottlenecks and large waiting times.

The reports were interpreted to see if the results confirmed that the identified bottlenecks were removed and that the optimized system version allows for increased productivity and improved “Takt time”. The results are presented in Figure 9 and Figure 10.

The results for the dynamic assembly line 1 (LAD 1) (eight workstations with eight operators (OP1–OP8)) show that the workstation for OP1 is idle for 6.33% and busy for 93.67%, the workstation for OP2 is idle for 6.65% and busy for 93.35%, the workstation for OP3 is idle for 6.80% and busy for 93.20%, the workstation for OP4 is idle for 6.97% and busy for 93.03%, the workstation for OP5 is idle for 6.65% and busy for 93.35%, the workstation for OP6 is idle for 6.97% and busy for 93.03%, the workstation for OP7 is idle for 7.21% and busy for 92.88%, and the workstation for OP8 is idle for 7.28% and busy for 92.72% of the total simulation time.

The results for OP1–OP8 are the same as the ones for OP26–OP33.

The results for rotary assembly line 1 (10 workstations with 10 operators (OP9–OP18)) show that the workstation for OP9 is idle for 6.33% and busy for 93.67%, the workstation for OP10 is idle for 6.56% and busy for 93.44%, the workstation for OP11 is idle for 6.73% and busy for 93.27%, the workstation for OP12 is idle for 7.05% and busy for 92.95%, the workstation for OP13 is idle for 6.49% and busy for 93.51%, the workstation for OP14 is idle for 6.65% and busy for 93.35%, the workstation for OP15 is idle for 6.96% and busy for 93.04%, the workstation for OP16 is idle for 6.49% and busy for 93.51%, the workstation for OP17 is idle for 6.65% and busy for 93.35%, and the workstation for OP18 is idle for 6.81% and busy for 93.19% of the total simulation time.

The results for OP9–OP18 are the same as the ones for OP34–OP43.

The results show that a productivity of 296 pieces could be obtained for two shifts. The productivity of 296 pieces from the simulation is lower than the estimated productivity of 320 pieces. Also, the “Takt time” from the simulation (for each of the two identical sides of the proposed structure of the assembly line) is 6.08 min, which is lower than the calculated “Takt time” of 5.63 min, but higher than the “Takt time” from the preliminary simulation (6.80 min).

Table 3. Comparison of results produced by WITNESS Horizon simulation between the initial and the proposed structure.

	Initial Structure Fixed Stations + Rotary Line	Proposed Structure Hybrid Assembly Line
Output Production (2 shifts—based on volume estimation)	160 [pcs]	320 [pcs]
Takt Time (based on volume estimation)	5.63 [min]	5.63 [min]
Blocked Time (range and distribution type)	varies between 1.67% and 9% (uneven distribution)	0%
Idle Time (range and distribution type)	varies between 5.17% and 14% (uneven distribution)	varies between 6.33% and 7.28% (even distribution)
Output Production (2 shifts—obtained after Witness Horizon simulation)	132 [pcs]	296 [pcs]
Takt Time (obtained after Witness Horizon simulation)	6.80 [min]	6.08 [min]

shows a comparison between the results obtained after simulation for the initial structure and the proposed structure of the assembly line.

The productivity of the proposed assembly line (one side) increased by 12% compared to the initial assembly line. The productivity of 296 pieces for the proposed assembly line is significantly increased compared to the productivity of 132 pieces for the initial assembly line. A capacity of 71,040 pieces per year could be obtained considering the productivity of 296 pieces for two shifts.

A “Takt time” of 5.63 min represents the ideal cycle time; however, in practice, variations exist relative to it, influenced by factors such as the following:

• Operator dexterity and differences in dexterity levels among operators;
• Idle times associated with various activities, such as retrieving components to be integrated into wiring harnesses, positioned behind the workstation (in the case of rotary assembly lines), and the time required for moving the trolleys from the end of the assembly line to dimensional/electrical/packaging control workstations [29];
• The unequal distribution of the number of operations among workstations or difficulties in equalizing the number of operations due to wiring harness architecture/component composition—for example, connector diversity/number of wires inserted in each connector, branches of different lengths, and different wire protections (depending on the area), which have different application times.

3.3. Validation of Simulations Using Simplex Algorithm

Information obtained previously with WITNESS Horizon Software is validated with the Simplex algorithm. The first three initial and final values for the three workstations are shown in

Table 4. Initial and final values for three workstations.

Initial Values		Final Values
Workstation	Time (min)	Workstation	Time (min)
1	5.5	1	5.30
2	5.5	2	5.41
3	5.6	3	5.43

.

The Simplex algorithm will be applied to the first three processes. The information is presented in

Table 5. Information about the first three processes.

Process/Workstation	W 1	W 2	W 3	Restriction
Process 1	1.9	1.9	1.9	5.30
Process 2	1.8	1.8	1.9	5.41
Process 3	1.8	1.8	1.8	5.43
	1.8	1.8	1.9
max[f(x)] = $1.8x_1+1.8x_2+1.9x_3$
$\left\{\begin{array}{c}1.9x_1+1.9x_2+1.9x_3\le 5.30\\ 1.8x_1+1.8x_2+1.9x_3\le 5.41\\ 1.8x_1+1.8x_2+1.8x_3\le 5.43\end{array}\right.$
$x_1, x_2,x_3\ge 0$
⇒ max[f(x)] = $1.8x_1+1.8x_2+1.9x_3+0s_1+0s_2+0s_3$
$\left\{\begin{array}{c}1.9x_1+1.9x_2+1.9x_3+s_1=5.30\\ 1.8x_1+1.8x_2+1.9x_3+s_2=5.41\\ 1.8x_1+1.8x_2+1.8x_3+s_3=5.43\end{array}\right.$
$x_1, x_2,x_3,s_1,s_2, s_3\ge 0$

.

Information from the previous system is centralized in

Table 6. Simplex table for the first three processes.

${\mathit{x}}_1$	${\mathit{x}}_2$	${\mathit{x}}_3$	s1	s2	s3	f	b	Ratio
$1.9$	$1.9$	$1.9$	$1$	0	0	0	$5.30$	2.80
$1.8$	$1.8$	$1.9$	0	1	0	0	5.41	2.85
$1.8$	$1.8$	$1.8$	0	0	1	0	5.43	3.02
−$1.8$	−$1.8$	−$1.9$	0	0	0	1

.

In the previous iteration, 1.9 will be the pivot.

Since the last row contains both zero and positive values, the iterations will be concluded, resulting in $x_1$ = 0,$x_2=0$, $x_3$ = 2.80. At this point, the optimal value is 5.32.

The same algorithm will be applied in the following workplaces to verify whether the final values obtained from the simulation are confirmed in

Table 7. The first iteration for the Simplex algorithm.

${\mathit{x}}_1$	${\mathit{x}}_2$	${\mathit{x}}_3$	s1	s2	s3	f	b	Ratio
1	1	$1$	$0.53$	0	0	0	2.80
$-0.1$	$-0.1$	$0$	−1	1	0	0	0.11
$0$	$0$	$0$	−0.95	0	1	0	0.41
0.1	0.1	0	1	0	0	1

.

On the last line are zero and positive values, thus, iterations will be completed. The same algorithm was applied to the following work points to analyze whether the final values obtained by simulation were confirmed. This information is presented in

Table 8. Initial values and final values for workstations 1–10.

Initial Values		Final Values
Workstation	Time (min)	Workstation	Time (min)
1	5.50	1	5.30
2	5.50	2	5.41
3	5.60	3	5.43
4	5.60	4	5.32
5	5.50	5	5.41
6	5.60	6	5.32
7	5.51	7	5.32
8	5.51	8	5.41
9	5.94	9	5.32
10	5.94	10	5.34

.

The solutions provided through WITNESS Horizon Software and supported by the Simplex algorithm meet the constraints of process optimization. The integrated approach in this article is efficient and converges to obtaining an optimal solution, which shows a beneficial robustness for the realized scenarios.

3.4. Methods for Production Waste Collection and Recycling in Wiring Harness Manufacturing

Due to the valuable metals they contain, recycling waste generated from wiring assembly is necessary for pollution reduction and the recovery of limited resources. Wiring harness manufacturers generally adopt a policy to reduce waste and streamline their collection. For wire scraps, connectors, protections, and other elements resulting from wiring production, selective collection is attempted wherever possible, based on the material type and the stage in the production process whereat the waste occurs. Collection is carried out at each stage using temporary storage containers. The final storage is facilitated through large containers, then transported to specialized companies engaged in waste recycling from various production sectors. Recycling firms are dedicated to developing new recycling methods to increase material recovery rates, particularly for valuable metals like copper, gold, and silver [24]. This involves reducing both the number of manual waste sorting operations and the environmental impacts of the recycling process, including carbon emissions.

Usually, the production waste collection is managed by integrating collection bins within a wiring harness assembly line. The most common collection systems include at least one of the following features:

• Collection bins or stations—At various points along the assembly line, collection bins are strategically placed to gather scrap materials such as cut-off wire lengths, excess insulation, and other discarded components;
• Segregation and labeling—When cycle time and available space are not constraints, the collection bins can be labeled to indicate the type of material that should be placed within them. For instance, there may be separate bins for copper wire scraps, plastic insulation scraps, and other specific materials;
• Centralized collection points—Depending on the size of the production facility, there may be centralized collection points where all the collected bins are gathered for later processing.

Manufacturers can efficiently manage scrap materials by integrating collection bins at strategic points along the assembly line, making reclaiming valuable resources and reducing waste easier. A typical example of a scrap collecting system in wiring harness production involves the organized collection and segregation of discarded or excess materials from the manufacturing process. Once the bins are filled or at regular intervals, the collected scrap materials are transported to recycling facilities or processing centers where they can be reclaimed, processed, and prepared for reuse in the production process.

The production waste collection system implemented in the proposed configuration of the assembly line consists of placing the collection bins along the line, as shown in Figure 11.

The collection bins are emptied regularly into centralized collection bins that are later prepared for transportation to recycling facilities.

Scrap recycling can be performed within an internal facility or by subcontracting to a specialized company. However, harness manufacturers have also been interested in developing new recycling methods that target production waste coming from wiring harness production.

For example, Yazaki Corporation patented a method for recycling waste wire harnesses that took the wiring recycling process a step further. They conceived a method that allows the collection and reuse of magnesium hydroxide. This research’s starting point was triggered by replacing vinyl chloride resin wire coating with non-halogen or low-halogen coating. The replacement was needed because it was discovered that vinyl chloride resin releases harmful hydrogen chloride gas, affecting human health. However, since the first usage of non-halogen or low-halogen fire-retardant resin composition as an electrical insulator and wire coating material, a new challenge has arisen regarding the recovery of the magnesium hydroxide, used as a fire retardant, that is found in the new resin type. A preparation step is required before applying the new recycling method, and the wire must be separated from connectors, clips, and protectors [30].

Many methods are applied in recycling wiring harness production waste; recovering as many elements as possible is mandatory for environmental protection and resource usage efficiency. Even though each recycling facility has its process and method, it can be observed that some phases of the recycling process tend to be common. The process mainly consists of removing and sorting the components of a wiring harness according to their material composition using different techniques, like cutting and shredding and applying different techniques to separate plastic from metal parts, such as electrostatic separation [31]. The main goal is to obtain materials with very similar technical characteristics to non-recycled materials, and to integrate them into the production cycle.

4. Conclusions

This research delves into the various factors influencing the logistical flow of wiring harnesses, and offers practical optimization solutions. These solutions are tailored to tackle the challenges posed by the flexible demands of the automotive market, making the research directly applicable to industry professionals, researchers, and stakeholders. The proposed hybrid structure, a pioneering concept in the industry, marries the adaptability and space efficiency achieved by streamlining and clearly defining workstations with the advantages of the rotary assembly line. This innovative approach is designed to boost production capacity without the need for facility expansion or subcontracting.

WITNESS Horizon software was used to model a virtual prototype of the assembly line to analyze performance indicators and identify necessary changes. The simulation of the initial architecture generated reports that helped pinpoint flow concentrators, which often lead to bottlenecks. A mathematical model calculated the workstation cycle times and proposed a new system architecture to optimize production. Modifications to the workstations resulted in a 12% increase in productivity and a reduction in “Takt time” from 6.80 to 6.08 min.

The proposed wiring harness assembly line has the potential to improve efficiency and flexibility; however, several limitations must be considered. The main concern is the initial investment required to reconfigure production facilities and train staff. Future research will focus on optimizing scalability, reducing costs through advanced automation, and enhancing adaptability to changing demands. Additionally, real-world pilot studies are essential to validate theoretical benefits and address unforeseen challenges.

In this context, the research also underscores the crucial importance of sustainable practices by integrating a production waste collection method within the proposed structure. The solution described not only demonstrates the potential for cost savings by efficiently collecting and recycling the scrap resulting from production, but also the significant environmental benefits of developing new methods for obtaining purer materials after recycling and integrating recycled materials into new products.