Study of the effects of complexity on the manufacturing sector

There are factors linked to the complexity of manufacturing systems that influence the development of processes and affect the indicators of a company. The objective of this work is to identify the effects and factors that generate complexity in an economic sector, additional management methods and indicators that allow a proposed solution. The methodological approach is based on the instrument developed by the University of Bayreuth, which allows the identification of effects, factors, methods and management indicators. The study is based on a sample of 71 small and medium enterprises from the city of Cartagena, Colombia. The results show the relevant factors of complexity according to their type and origin that intervene in the manufacturing sector, as well as the different methodologies and management indicators that allow managers to improve the performance of their companies and the sector


Introduction
Nowadays, companies must adapt to continuous changes and be competitive in order to stay in the market. As a strategy they seek to increase productivity, improve processes, performance and increase customer satisfaction. But there are factors such as complexity. This is linked to the high number of variables and uncertainty. According to [1], complexity derives from structural aspects within the system being established by the variety of elements that make it up and the relationships that intervene. It should be noted that in a manufacturing system associated production factors act, (i) with the materials when they do not meet specifications of time, quantity and quality, (ii) with labor when there are changes in the pace of work, absenteeism and accidents, (iii) with the machines when they fail, absence of spare parts and tools. Where variables of quantity and variety of processes, products and services make systems unstable and complex.
As a consequence of this problem, there are imbalances in the plans and reprogramming, which increase the costs reflected in the consumption of hours of preparation of the machines, repair times, overtime, among others. Understanding the existence of these characteristics allows us to consider that manufacturing companies are complex systems in complex environments and that administration or management becomes complex. Often the studies that have been developed are more and more frequent due to the eagerness of organizations to find a complex education of processes minimizing costs, increasing revenues and improving competitiveness in local and international markets [2][3][4][5].
According to a previous analysis in a review of the literature it could be inferred that there is a weakness or scarcity of such management in the manufacturing sector. Given this situation, a country like Germany is considered a pioneer in the management of complexity analysis techniques in the manufacturing sector, using methodologies developed by Bayreuth University [6,7]. Consequent to this [8] in Sweden they carried out a comparative study taking as a sample three companies of the sector. Similarly, in Mexico, taking into account the previous studies mentioned above, an analysis was carried out on companies in the manufacturing sector, more specifically in the area of metalworking [9]. Based on these antecedents, the purpose of this research is to try to contribute to the diffusion of a methodological approach of management of complexity in companies of the manufacturing sector of the city of Cartagena-Colombia. The study required an analysis of different economic sub-sectors, considering characteristics of structure, variety and relationships. The results obtained support managers and administrators in the decision-making process, showing an approach to managing the complexity of manufacturing processes by identifying the relevant associated factors that affect performance indicators. This work is constituted in the following way: Sect. 2 presents the theoretical context of complexity in manufacturing systems. Section 3 describes the methodology addressed. Section 4. The results obtained. Section 5 provides the conclusions. Finally, acknowledgements.

Theoretical foundation
The complexity of a manufacturing system is related to several variables such as origin, quantity, variety, time and system relationships. Regarding the origin, for [10] the reasons that cause complexity can originate from inside or outside the company and can be generally classified in three main categories: internal, external and total. Internal complexity is associated with flows within the manufacturer, and can be caused by factors external to the organization called external complexity, which is associated with flows to suppliers or customers. And total complexity covers all internal and external complexity (see Fig. 1).
According to time and its behaviour, according to [11] complexity in manufacturing systems can be static or dynamic. Static complexity refers to a characteristic that can be associated with the systems, and also with the production processes, aligned with the structure of the facilities or the structure of the plant and considers the degree of difficulty in its management and control. Such complexity becomes important when studying the possible design of an installation or plant. And dynamic complexity refers to the analysis of systems over time, in other words, it studies the trend of the real states that the process assumes within the time considered (see Fig. 2).
In manufacturing systems, it is necessary to manage the complexity of the process. To do this, the relevant factors that affect performance indicators must be identified. According to [12] there are determining factors in the complexity of manufacturing such as: (i) the structure of the product, that is, the number of different articles and for each product: number and type of sub-assemblies, cycle times, lot sizes, type and sequence of resources required to produce it; (ii) the structure of the plant, i.e. the amount and types of resources, design, setup times, maintenance tasks, downtime, performance measures; (iii) the planning and scheduling functions, with three components such as planning and scheduling strategies, the number, content, timing and priority of documents used for planning and scheduling, and the decision-making process; (iv) the flow of information during the decision-making process, teamwork, within the plant with other departments and externally with other plants, suppliers and customers; (v) the dynamism, variability and uncertainty of the environment such as customer changes, breakdowns, absenteeism, inaccuracy of data and unreliability; (vi) other functions within the organisation such as training, political information. As regards static complexity, it has been studied that it has a negative effect on productivity and quality [13], since a high number of products and/or the variety of their components generates difficulties in the design and operation of assembly lines [14]. Likewise, the static complexity related to the product has a negative impact on production costs [15]. On the other hand, dynamic complexity also has a negative effect on costs, so the higher the dynamic complexity, the higher the costs [16].
Experts say there are three main strategies to combat the negative effects of complexity: avoiding it, reducing it or controlling it. Studies by [17] claim that 25% of total costs for manufacturing companies are due to product and process related complexity. In short, companies tend to use strategies to avoid, reduce or control complexity. According to [18], the use of tools and technologies to support complexity management is widely used and recognized.

Method
Methodologically, six stages are needed to carry out the research: (i) Study of the manufacturing sector, (ii) Sectoral analysis of complexity, (iii) Effects of complexity, (iv) Improvement methods, (v) Complexity factors and (vi) management indicators (see Fig. 3).
Initially a study of the manufacturing sector in the city of Cartagena-Colombia is developed. This work derives from the participation of one of its authors in a research project and initiative of the Cartagena Chamber of Commerce in order to strengthen the development of the productive, managerial and associative capacities of Colombian PyMES belonging to the city's clusters. In this phase, a general analysis is made of small and medium-sized enterprises (PyMES) in the manufacturing industry, taking into account their classification according to different economic activity, number of companies that make them up and their total assets. As a second phase, the complexity analysis methodology is applied to the prioritized sub-sectors by means of questionnaires and analysis of company data. In the third phase, the effects generated in the sector are analyzed, based on the elements that intervene in a system such as plant, process, product and planning. Finally, the factors are identified taking into account their classification according to type and origin, then the improvement methods and performance indicators are determined to be able to manage the complexity that affects the variability of the system, in this last part, the methodology used by [6,7].
It should be noted that in order to develop a methodology for quantifying the determinants of complexity and effects, the formal description of complexity, which remains very abstract, requires first the identification of theoretically possible determinants of complexity and then the selection of determinants of complexity that are relevant in practice [7].

Manufacturing sector study
The industrial environment of different countries is classified between large, medium and small industry. This depends on various parameters that are used to establish the differences. These include gross sales, number of employees, total assets, tax rates or other economic mechanisms or formulas that attempt to facilitate this classification [19].
In Colombia, in accordance with Law No. 905 of 2 August 2004, companies are classified according to the number of staff employed and the total assets expressed in terms of current legal monthly minimum wages (see Table 1).
According to the Cartagena Chamber of Commerce, there are currently 32535 companies distributed as follows: 29576 (91%) microenterprises, 2792 (8.6%) PyMES and large companies 167 (0.5%). Selecting the manufacturing industry, this is represented by a total of 3982 companies equivalent to 12%, until the end of 2019, the number of microenterprises were 3669, PyMES of 278 divided into 224 as small and 54 medium and in the case of large companies a number of 35. In terms of percentages 92% is equivalent to microenterprises, 7% to PyMES and 1% to large industries (See Table 2).

Sectoral analysis of complexity
In the first instance, the companies in each economic subsector (population) were identified and those representing a 70% accumulated participation were chosen, which is taken as a sample criterion to be able to develop the methodology. This means that 71 out of a total of 128 firms were selected, representing 55.5% of the total size. Figure 4 shows the number and names of the companies selected by subsector; (F) Food, (M) Metalworking, (C) Chemical, (P) Plastic and (L) Lithographic.
According to [7] In order to develop a methodology for quantifying complexity determinants and effects, the formal description of complexity, which remains very abstract, requires first the identification of theoretically possible complexity determinants and then the selection of complexity determinants that are relevant in practice.
The study is based on the complexity analysis methodology developed by the University of Bayreuth as part of a research project funded by the German Federal Ministry of Economics and Technology, which consists of determining the effects of complexity on production processes in order to convert them into a model. The configuration of the methodology addressed helps to manage complexity in manufacturing companies and allows the control of complexity in the process .1 According to the methodology, in stage 1 the information about the company is obtained, from a structured questionnaire, then it continues with stage 2 where the effects of complexity are listed, consequently in stage 3 the set of suitable methods to control the effects of complexity is shown and finally in stage 4 the drivers of complexity "Drivers" and key performance indicators ("KPI") are related (See Fig. 5).
As part of the methodology described in the framework of the project developed by the University of Bayreuth, a series of diagnostic tools were applied, such as interviews and surveys to a large population in the industrial sector, where with the help of experts it was possible to identify the effects attributed to greater complexity, thus providing a total of 50 different effects. Similarly, more than 4800 combinations of each complexity factor, each target field and each dimension were quantified, examined and evaluated using matrices. In this way, each of the 50 effects is compared with 44 methods and instruments, resulting in a total of 2,200 documented combinations in a two-dimensional matrix. The results are used to finally develop a web-based configuration tool that allows manufacturing companies to find out which optimisation methods or tools adapted to their company's indicators are suitable for managing the effects of complexity in their manufacturing process.
Given the above, taking into account that in this research the analysis is of a sectoral nature. The information was generated through field visits and the application of structured surveys, directed at production managers and directors, adjusted to characteristics such as (i) Type of production process, (ii) Number of employees in production, (iii)  The results show that 61% of the companies work under a type of make-to-order process, 58% have between 11 and 50 employees in the production area. 51% of the companies handle between 11 and 50 references of different products. 55% have more than 50 customers. 82% agree that they have between 1 and 10 work stations for the development of the production process and 56% of the companies have a type of semi-automatic operation. Figure 4 shows the number and names of the companies selected by subsector (See Table 4).

Effects of complexity
To obtain the effects of complexity, a study was carried out in each of the companies (71 companies in total) by subsectors, using an interview technique and surveys with qualified and experienced personnel, considering relevant aspects in the five elements that make up a production system (i) Plant, (ii) Process, (iii) Products, (iv) Persons and (v) Planning. Given the above, (14) effects were selected that occur most frequently in the different sectors, which represent 70% of the total (See Table 5).
To obtain the effects of complexity within the instrument, the criteria of the methodology described above were used as a basis. It proposes a checklist of fifty (50) effects. The Table 6 shows the selected effects which will provide the relevant information for the following stages.

Complexity factors
The complexity factors of a manufacturing system can be of an internal type which are generated by the decisions and elements within the organization such as the process, the product, the planning, the people and the plant. And  the external factors linked to the relationships between suppliers, customers, market, competition and regulations. In order to be able to determine the factors of complexity in a manufacturing system. It should be noted that the instrument applied determines the relevant factors of each of the effects, some of which are duplicated. Table 7 shows a grouped and integral relationship for all the economic sectors taken into account in the study. Taking into account the classification of complexity from its origin and type.

Improvement methods
The tool proposes a checklist of forty-four (44) improvement methods or tools. Through the use of a configurator that generates methods from the identified effects, it is possible to identify mechanisms that make it possible to counteract, reduce and eliminate complexity in manufacturing systems (see Table 8).
As can be seen in Table 8, a total of 36 methods out of 44 need to be implemented in the different sub-sectors such as food (F), metalworking (M), chemical (C), woodworking (W), plastic (P) and lithography (L). It is clear that these sub-sectors share common methods for tackling complex problems.

Complexity management indicators
In addition to identifying complexity factors, it is necessary to measure it in order to relate it to costs and other performance indicators in order to establish opportunities for improvement [20]. Table 9 shows the end result that provides the opportunity to establish a complexity management system based on key indicators that minimize the effects that occur in the manufacturing sector in a comprehensive manner.

Conclusion
This research was based on the complexity analysis methodology developed by the University of Bayreuth, unlike the instrument applied to an integral sector. The study was developed in the manufacturing sector in the city of Cartagena-Colombia, taking as a sample the small and medium enterprises (PyMES) in the manufacturing industry, by their classification according to the different economic activity, number of companies that make them up and their total assets. From this, the objective was achieved of being able to identify the most representative effects in each of the subsectors and the elements with the greatest impact on the manufacturing sector, leading to the identification of complexity factors by origin and type. As a result of this, the generating instrument made it possible to highlight the Part of disposal areas The part of disposal areas is determined by dividing the disposal areas by the total production area. The unit of the KPI is percent.
To reach a good result the part of disposal areas should be quite small (Disposal areas / Total production area)*100 Process efficiency The efficiency of the process is determined by dividing the value-added work by the total work spent. Work can also be measured in a unit of time. The unit of KPI is the percentage. To achieve a good result the process efficiency must be quite high (Value adding work/total spent work) * 100% Rush order rate (production and purchase) The rush order rate is determined by dividing the number of rush orders by the total number of orders. The KPI's unit is percent. To reach a good result the rush order rate should be quite small (Number of rush orders/total number of orders)*100%

Scrap rate
The rejection rate is determined by dividing the defective production by the total production. The scrap rate is determined by dividing the scrap by the total production. The unit of both KPIs is the percentage. To achieve a good result the share of defective production must be quite small (Scrap/total production) feedback8% Training efforts per employee and year The employee training effort is determined by dividing the total costs of the employee training measure by the total number of employees. The unit of the KPI is the euro/employee. To achieve a good result, the training effort of the employees must be quite high Total costs of employee training measure/total number of employees different methodologies in an integral manner that are necessary to be implemented in the different subsectors. Finally, it is possible to establish a complexity management system based on key indicators that minimize the effects produced in the manufacturing sector. This work allowed to reduce the gap between the existing theoretical foundation and the applied practical projects, since it had not been considered before by the companies of the manufacturing sector.

Future work
For future research it would be useful to apply this methodology to another type of economic sector or make an analysis of complexity in the manufacturing sector taking into account the different processes types, such as flow, workshop type or per project.