A proposal for the supply chain design: A digitization approach

The logistics network of an automotive company in Mexico, was analyzed to propose a better logistics network in the country to improve delivery times to customers. The network design considers elements of digitization of Greenfield Analysis and Network Optimization processes. Taking into account the information given by the company, it was possible to obtain optimal scenarios for better operation, which involved the construction of distribution centers throughout Mexico. Received on 11 April 2020; accepted on 16 April 2020; published on 27 April 2020


Introduction
The globalization level reached in today's World must go in hand with an infrastructure that is capable of satisfying all markets. In this case, we are talking about Supply Chain [1]. This topic has evolved by using tools to face the most challenging organizational problems [2]. Supply Chain global companies face many more challenges tan small supply chain companies since there are more and more complex planning and control variables [3][4][5]. For example, they need to consider sustainability, balance economics, environmental care, and social issues, such as creating strong job positions for the population [6]. Some companies can invest resources for including in their supply chain, environment-friendly steps such as renewable energies, or recycling processes [3]. Today, organizations seek for all these concepts because they know they add value to their brands.
However, they must also control the production process, to achieve this, organizations must also

Literature review
Top executives must understand this planning process for transmitting it to the whole organization, and this means that the proper management of each supply chain link will allow the smooth operation of the methods. Even selecting suppliers should be as per the objectives of the organization so to meet the established timing and quality standards as well, and for them to adjust appropriately. Let's say, for example; they should meet the minimum delivery amounts or a given period of time; this way, the organization assures smooth logistics.
The sustainability concept has been increasing during the last years. Besides the concern of having proper logistics planning, the whole supply chain should include sustainability in its processes to increasingly reduce the manufacturing CO2 emissions. It is clear that polluting emissions do not come only from manufacturing processes but also from means of transportation, and that is why the transportation network within the supply chain must include a strategy aimed at controlling polluting emissions [7]. There must be multi-disciplinary teams forming new supply chains since the economic, financial, environmental, and social aspects must be taken into consideration.
There must be an optimal product moving process from the outset point up to the sales point: this is, the planned logistics must meet in good time and proper form the production of goods, as well as the delivery thereof.
To achieve proper production planning, it is critical to developing demand forecasts, so this way, you can control production with the right amount of waw material, equipment, workforce, etc. These variables must be calculated appropriately since they highly impact on the organizations expenses, and not having a clear view of the demand forecast might lead the company to a shady path.
There are models developed before 1990 for supporting supply chain management that analyzed and provided the organizations with the best expense, tax, and fare-minimization options. Later on, models were exchange-rate variability-focused [8]. Therefore, since it is an international business, the economy has a significant role to play because it impacts the supply chain prices, costs, and expenses in the production Country, as well as in those Countries where organizations want to export their products to. There are some models providing customers with this analysis, such a Hodder and Dincer, which includes the financial impact derived from the Government subsidy and tax reduction, thus allowing a correlation between the changes in prices within the global markets [8].
There are even other models such as the Breitman and Lucas which allows studying the manufacturing plant location, the production amount, the required quantity of raw material, and the introduction of new products into the market.
Supply chain models need to involve more than the geographic location of manufacturing plants, but also their internal and external flow, as well as optimum suppliers who meet the sustainability organization's objectives, and the internal handling of material in each process [9].
Supply chains must expand their variables to respond to the needs of today's globalization in which we are immersed; this is the reason why they need to address alternative, trustable objectives flexibly by calculating operation costs and assets [9, 10]. Also, the ideal thing is there were multi-level production and distribution analysis due to the large amount of data required for analyzing supply chains. Many current models just address the supply chain first stage, neglecting other steps, such as performance and supplier selection [11][12][13]. Most companies seek for proper supply chain planning, finding accuracy and ethical management in each stage for preventing unnecessary expenses, avoidable delays, etc.; this is because, more often than not, companies don't own the facilities where they produce. Therefore, the supply chain mus control the variables involved must; for example, determine the minimum order quantities per supplier, plus budgetary restrictions, number of suppliers, geographical preferences, and capacities. Besides, these models should have objectives and constraints for assessing the impact quality, lead times, and service level. This last concept, service level, is one of the organizations' most wanted. Customer satisfaction is guaranteed by having a proper supply chain.

Supply Chain Design
For this study the data used were scaled by a factor to not disclose confidential company information. Therefore, the information presented does not correspond to reserved information.
For this project, 50,843.00 data on vehicle demand; were collected; which were divided into 3 different origins: Non-North America (Product 1, Product 2, Product 3, Product 4, Product 5), North America (Product 6, Product 7, Product 8, Product 9, Product 10, Product 11, Product 12, Product 13, Product 14, Product 15 and Product 16) and Mexico) Product 17, Product 18, Product 19 and Product 20). This project will only focus on vehicles from Non-North American origin; which reduces the database to 25,696.00 vehicles, with 5 lines of vehicles divided in the following way (see Table 1): 2 EAI Endorsed Transactions on Energy Web Online First For the analysis, the following information ( Table  2) provided by the company per vehicle unit was considered: The information provided goes from january 2018 to October 2018, from which the monthly historical demand by distributor was obtained.

Percentage of allocation of sales by city
When performing an analysis of the database it was obtained that 79% of sales of vehicles was held in 21 cities and the 21% remaining is divided into the other 42 cities (See Table 3 and Figure 1).

Assignment of cases
Based on the concern of the automotive company, the cases were classified as follows: Success: It is considered success when units were ordered and sold at the same distributor in less than 45 days • Error #1: Units that were ordered and sold at the same distributor, but the 45 days credit were overcome. • Error #3: Units that were ordered and sold at different distributors in the same city, but the 45 days credit were overcome.
• Error #4: Units that were ordered and sold at different distributors in different cities, but not exceeding from the 45 days of credit.
• Error #5: Units that were ordered and sold at different distributors in different cities, but the 45 days of credit were overcome.
Errors 3, 4 and 5 are considered as the most serious, since they involve overcoming the 45 days of credit and some kind of exchange (see Table 4): Table 4, it can be further enhanced that 32.37% of the units are ordered and sold at the same distributor, without exceeding the 45 days credit. In the same way, it stands out that the most disturbing errors (3, 4 and 5) add up to 28.68%, which could be dramatically reduced if there was a centralized distribution facility, with an optimal location and distribution network. By using the cost information provided by automotive company and the information obtained in the previous • The cumulative percentage of error 4 and 5 = 26.25% • The average transport cost ($) = $141 By eliminating such errors, the automotive company and its distributors might save approximately $2,516,195.76 usd. It is believed that by having a centralized distribution facility the other types of errors will decrease since distributors will have a higher turnover of inventory.

Sales forecast
The database was used to obtain the percentage of allocation of vehicles from each of the distributors and with the prognosis of vehicle sales, the number of units by vehicle line estimated for each of the distributors to sell through 2019 was obtained (Table 5 and Table  6). The table in percentage of allocation is presented  in table 3 and table 7 Information provided by the automotive company.

Additional information
Time in Product 5: Since the billing and delivery to the distributor dates are available, the average time in days that the automotive company takes to deliver a vehicle in their different points of sale throughout the country was calculated. The average number of days in Product 5 from wholesale billing to delivery to the distributor was calculated in 8.6 days. This retrieved data will be used as a before and after comparison with the results of the Greenfield Analysis and Network Optimization.
Average sale days: With the analysis performed it was verified that the average number of days that it takes to sell vehicle is 101 days, this data meets the information given by the automotive company.

Mathematical Model
The following analysis aims to define the ideal location for the new production plant, taking into account the place where each of the customers is located and their demands. It is based on the following mathematical method.
Where: d i x = x coordinate of the locality i d i y = y coordinate of the locality i w i = volumen for the locality i One of the main features of AnyLogistix (ALX) is that it takes into account actual strokes of distances from one point to another. In the same way, it considers suitable transport roads. By using the mathematical method, only coordinates are considered, but not actual distances or roads [14].

Software Used
AnyLogistix (ALX) is a specialized software that serves to address a wide range of supply chain management (SCM) issues, see figure 2. Decision making in supply chain management involves the use of quantitative methods, which normally are typically based on optimization or simulation. ALX allows finding appropriate locations and characteristics of the site through the optimization of the network, taking into account demands, capabilities, seasonality, types of products, customer locations, roads, fixed costs and variable and inventories. The above helps to make an informed decision based on costs, revenue, service levels, utilization rates and other parameters used in supply chain.
On the other hand, an analysis on how efficient the current distribution network is may be performed and verify if it is reasonable to reset, by opening additional stores or closing any existing. A simulation model of ALX allows capturing the randomness in the behavior 4 EAI Endorsed Transactions on Energy Web Online First  In a simulation model, the effect that each variable has within the system may be seen; which gives the possibility to dynamically change parameters during the experiment and to analyze their impact. A simulation does not produce an optimal solution from different combinations, but allows to study in a comprehensive manner particular dynamic scenarios and interdependencies of the supply chain. If more details are considered, then more opportunities to find improvements are created.

Location analysis
Several Greenfield Analysis scenarios were performed to obtain the optimal location for the distribution centers, which required the following information.
Eighty three distributors and one supplier were taken into account, which in this case is the Port of Veracruz, since those are Non-North America vehicles. Figures 5, 6 and 7.

Historical data
As enough data were available, it was decided to use the historical demand from January to September, since the information from October did not cover the entire month and also, some of the sales information of the distributors was missing.

Population index
One of the advantages of the software is that it has an option to perform experiments with population density. Therefore, by conducting a comparison between the population data from INEGI and the figure 1 and 4 of vehicle sales by state, it is possible to observe that for greater population densities, there will be higher vehicle sales.

EAI Endorsed Transactions on Energy Web
Online First

Products
As mentioned above, the following lines of vehicles will be used:

Results
The following results were obtained: Since the software has geolocation, it is possible to locate the exact coordinates from the distributors. By using the database above mentioned, the program gave a result of a distribution center at latitude 19.96 and longitude -98.29 coordinates, which correspond to the state of Hidalgo Figure 8 shows the distributors, the located distribution center and the provider that is the Port of Veracruz. [

Conclusions
After analyzing the case study shown in this work, the digitization process of the supply chain is shown. The use of automated tools that allow the development of digital twins such as ALX, work as a support in the decision making of network design. Enablers of the digital supply chain such as agility, flexibility, real-time data exchange, are used to solve the problem of the company under study.