topic simulation of kinh dos fresh bread supply chain in ninh thuan province

HOCHIMINH CITY INTERNATIONAL UNIVERSITYSCHOOL OF INDUSTRIAL ENGINEERING & MANAGEMENTTOPIC: SIMULATION OF KINH DO’S FRESHBREAD SUPPLY CHAIN IN NINH THUANGROUP: Group 37 Advisor: Trần Đức

HOCHIMINH CITY INTERNATIONAL UNIVERSITY

SCHOOL OF INDUSTRIAL ENGINEERING & MANAGEMENT

TOPIC: SIMULATION OF KINH DO’S FRESH

BREAD SUPPLY CHAIN IN NINH THUAN

PROVINCE

GROUP: Group 37 Advisor: Trần Đức Vĩ

List of Full name - ID:

Trần Tấn Phát - IELSIU22224

Huỳnh Khánh Nguyên - IELSIU22074

Phạm Bảo Minh - IELSIU22188

Vũ Hoàng Duy - IELSIU22179

Ho Chi Minh city, Vietnam

August/2023

TABLE OF CONTENTS

ACKNOWLEDGEMENT

ABSTRACT

LIST OF TABLES

LIST OF FIGURES

CHAPTER 1: INTRODUCTION

1 1 System requirement

1.1.1 Rationale

1.1.2 Problem statement

1.1.3 Introduction to Anylogistix

1.2 Objective

1.3 Scope and limitation

1.3.1 Scope

1.3.2 Limitation

CHAPTER 2: SYSTEM REQUIREMENT

2.1 Conceptual design

2.1.1 Idea

2.1.2 General flowchart

2.1.3 Detailed flowchart

2.1.3.1 Inbound process

2.1.3.2 Outbound process

2.2 Anylogistic model

2.2.1 Input

2.2.1.1 Data collection method

2.2.1.2 Data collection

a Customers (Retailers)

b DC and factory

c Products

d Demand

e Facility expenses

f Inventory

g Vehicles

h Paths

i Processing cost

j Processing time

l Shipping

m Sourcing

2.2.2 Process and output

2.2.3 Validation

CHAP 3: EXPERIMENTAL DESIGN

3.1 Scenario 1: Add one more DC

3.1.1 Process and output

3.1.1.1 GFA experiment

3.1.1.2 Simulation experiment

3.1.2 Result of Scenario 1

3.2 Scenario 2: Add 2 more DCs

3.2.1 Operation description

3.2.1.1 GFA experiment

3.2.1.2 Simulation experiment

3.3 Final result and conclusion

CHAPTER 4: CONCLUSION

4.1 Advantages and Disadvantages

4.1.1 Advantages

4.1.2 Disadvantages

4.2 Application

REFERENCES

APPENDIX

Appendix 1: Customers’ locations

Appendix 2: Customers’ demand

Appendix 3: Contribution

To complete the project with the topic: “Simulation of Kinh Đô bread supply chain inNinh Thuan Province”, it is impossible not to mention the enthusiastic help of teachersand teaching assistants We extend our sincerest thanks to:

- Dr Trần Đức Vĩ, advisor, who provided our group with foundationalknowledge and enthusiastic courage in the course of accomplishing the report

- Miss Đoàn Thúy Nhã, the teaching assistant, who helped us with technologicalproblems and assessment during the experimentation in the simulation app -Anylogistix

We also want to extend our appreciation to lecturers and teaching assistants fromIndustrial Engineering Management for providing us with an opportunity to assess thistool and give us a clearer insight into our major

Mondelez Kinh Do company (or Kinh Do) is one of the most reputable companiesfamous for its traditional Vietnamese cake products With a nationally - widespreadsupply chain network, Kinh Do has been the first choice of Vietnamese customers fornearly 30 years This report shows the simulation of Kinh Do supply chain in NinhThuan province in the whole year 2022 with the support of the Anylogistixapplication The report is based on data derived from reliable sources as well as soundassumptions First of all, the existing supply chain is simulated with actual data as abase case Then, several hypotheses that are supposed to be able to optimize the profitare simulated as separate scenarios The comparison between the results of three cases

of simulation provides insights into the existing supply chain with its complexcomponents, as well as the potential of optimizing the profit for the business

LIST OF TABLES

scenario 1 and scenario 2

LIST OF FIGURES

Figure 4 Customers, DC, factory locations 19

Figure 6 Base case - simulation result (1) 28

Figure 7 Base case - simulation result (2) 29

Figure 8 Base case - simulation result (3) 29

Figure 9 Base case - simulation result - profit and lost 30

Figure 10 Base case - Revenue and Profit graph 31

Figure 12 Scenario 1 - GFA setting - period 33

Figure 13 Scenario 1 - GFA setting - number of sites 33

Figure 14 Scenario 1 - GFA setting - default constraints 34

Figure 15 Scenario 1 - GFA setting - unit setting 34

Figure 16 Scenario 1 -GFA result - New DC location

Figure 20 Scenario 1 - Simulation experiment - group 36

Figure 21 Scenario 1 - Simulation experiment - DCs

and Factory

36

Figure 22 Scenario 1 - Simulation experiment - Facility 37

Figure 23 Scenario 1 Simulation experiment

-Inventory

37

Figure 24 Scenario 1 - Simulation experiment - Paths 38

Figure 25 Scenario 1 Simulation experiment

Figure 29 Scenario 1 - Simulation result - Dashboard 40

Figure 30 Scenario 1 - Simulation result - Profit and

Figure 32 Scenario 2 - GFA setting - experiment

duration

42

Figure 33 Scenario 2 - GFA setting - number of sites 42

Figure 34 Scenario 2 - GFA setting - default constraints 43

Figure 35 Scenario 2 - GFA setting - unit setting 43

Figure 36 Scenario 2 - GFA result - New 2 DCs

Figure 38 Scenario 2 - GFA result - GFA DC locations 44

Figure 39 Scenario 2 - GFA result - GFA DC 2

Figure 42 Scenario 2 - Simulation experiment - DCs

Figure 50 Scenario 2 - Simulation results - Dashboard 49

Figure 51 Scenario 2 - Simulation results - Profit and

loss

50

Figure 52 Scenario 2 - Revenue and Profit graph 50

The purpose of the logistic simulation is to evaluate and optimize the efficiency,effectiveness, and cost-effectiveness of different logistical strategies, policies, andoperational scenarios By running simulations, logistics professionals can test different

"what-if" scenarios, assess the impact of potential changes, and make informeddecisions to improve the performance of their logistics operations

1.1.2 Problem statement

With a population of 99,959,381 people in 2023[1], Vietnam ranks 15th in theworldwide population ranking and has become a promising market for the foodindustry The fresh bread market witnessed the same trend with an increasing number

of competitors in its market share, namely Kinh Đô, Staff, Otto, Orion, etc Althoughgaining advantages from the blooming market, these companies simultaneously faceseveral challenges The average expiry day for instant food like fresh bread isgenerally 3 to 7 days [2] This short life cycle can put substantial time pressure on thefresh bread supply chain Insufficient facilities and distribution systems in Vietnamare also additional hindrances to meeting wholly customer demand We arerecognizant of unsatisfactory service levels and unoptimized costs, which come fromunelaborated supply planning, inaccurate market prediction and lack of cost-effectiveness in transportation

Kinh Do Corporation is a well-known food-and-beverage company in Vietnam thatproduces a variety of food-and-beverage products In 2015, Kinh Do sold 80% of itsshares to Mondelēz International, a multinational corporation based in the UnitedStates, which represents a significant example of cross-border investment andcollaboration in the food industry For the sake of simplification, we use the title

“Kinh Do” to refer to Mondelez Kinh Do Company in this report

Kinh Do has a complex supply chain with an extensive distribution network acrossVietnam and other Southeast Asian countries, as well as its sourcing of raw materials

from various suppliers Bread is a staple food in Vietnam and an important part ofKinh Do's product portfolio, making it a key area of focus for the company's supplychain management

Through our study of Kinh Do fresh bread in Ninh Thuan province specifically, wehave observed that there is only one distribution centre in Ninh Thuan provinceserving over 100 customers Being aware of some undiscovered shortcomings in theprovincial supply chain, we conduct a study on Kinh Do's bread supply chain Bysimulating different supply chain scenarios with varying numbers of distributioncentres, we aim to provide insights into the potential benefits and drawbacks of suchchanges and offer informed recommendations for optimizing Kinh Do's supply chain

in Ninh Thuan province

1.1.3 Introduction to Anylogistix

AnyLogistix is a supply chain optimization and simulation software developed byAnyLogic It enables users to model and analyze complex supply chain networks,evaluate strategies, and make data-driven decisions With AnyLogistix, users cancreate detailed models of their supply chain, simulate various scenarios, optimize keyperformance indicators, and visualize results It offers advanced optimizationcapabilities and helps businesses improve efficiency, reduce costs, and meet customerdemands effectively

1.2 Objective

By employing the AnyLogistix software for the supply chain of Kinh Do bread, ourobjective is to simulate the Kinh Do bread supply chain in Ninh Thuan Province tostudy the real-life system without disrupting the existing operations Furthermore, ourgoal is to assess the performance of the current supply chain, identify any deficiencies,and propose suitable solutions to optimize profitability and service levels

1.3 Scope and limitation

1.3.1 Scope

We will study the supply chain of Kinh Do fresh bread from 1/1/2022 to 31/12/2022.Our supply chain starts from a factory in Binh Duong Province to a distribution centre(with its warehouse) in Ninh Thuan Province and 100 retailers in Ninh Thuan

Province are the last destination for our products We focus on the demand for threesignature products: “Sandwich chà bông”, “Burger bò” and “Pizza xúc xích” due tothe limited information provided Also, it is reported by salers that these threeproducts earn significant revenue compared to others The evidence can be seen in

Another significant challenge was the accuracy of certain data Despite our bestefforts, crucial data points such as demand and transportation cost remainedunavailable Moreover, the limited scope of the bread supply chain project, confined

to a single province, represents an inherent weakness This restricts thegeneralizability and broader applicability of our results, as they may not capture thecomplexities of larger-scale supply chains

Despite these constraints, our group made efforts to mitigate these issues and derivedmeaningful insights within our available resources and expertise

CHAPTER 2: SYSTEM REQUIREMENT

2.1 Conceptual design

2.1.1 Idea

Demand for our products is recorded in the form of orders by the sales sector in NinhThuan Based on the inventory amount of the distribution centre, the manager decides

to make orders from the factory in Binh Duong With make- to-order strategy, thisfactory manufactures according to orders Then, the finished products are transported

by trucks to the distribution centre and come to retailers by vans

in creating goods in the factory In the early stage, the staff checks for available stockswhich will be transported to the warehouse if it is accessible Then, in order to makethe process continuous, they buy materials from reliable suppliers Here, they need toinspect carefully before combining and turning them into products on demand.Finally, things will be shipped from the production to the warehouse and finally to themerchants Following that, stores will sell things to customers, and the process will berepeated

Figure 1: General flowchart

2.1.3 Detailed flowchart

2.1.3.1 Inbound process

Inbound process happens at the factory when orders are received from DC In thebeginning, the personnel will first identify the product orders from the customers(particularly DC) before checking the adequacy of stock to see if the facility has

enough raw materials for production In case the available stocks are adequate, theprocess will go to the next phase, which involves sorting and getting ready the rawmaterials for manufacturing Followingly, the products will go through a qualityassurance process The qualified products are then packaged, labelled, and put ontotrucks for delivery to the DC By contrast, unqualified products are cancelled On theother hand, additional procurement is necessary if initial stocks are insufficient, andthey must then be examined for quality The unqualified materials will be returned tothe suppliers and the inventory turns back to the inadequate state, while the qualifiedmaterials will be sent to the sorting and preparation phase The detailed inboundflowchart is illustrated in the following figure:

Figure 2: Detailed inbound flowchart

other hand, if the supplies are not available, the staff will make orders to the factory.When receiving and unloading products, officers assess product quality Qualifiedproducts are then collected by orders and prepped for distribution, whereasunqualified products turn the process back to procurement activities The detailedoutbound flowchart is illustrated in the following figure:

Figure 3: Detailed outbound flowchart

2.2 Anylogistic model

2.2.1 Input

2.2.1.1 Data collection method

Choosing this topic, our team can receive several real relevant but unpublicizedstatistics from Kinh Do Company, which play an integral role in visualizing the pre-existing supply chain with high feasibility However, as mentioned before, some ofthe important data are inaccessible to us, so we have to base our assumed statistics on

analysis of articles, reports and documentation from reliable sources which are listed

a non-commercial edition of Anylogistic, we have to choose only 100 over 148customers for data input The locations of these customers are inside Ninh Thuan

Province For further detail, read Appendix 1

Figure 4: Customers, DC and factory locations.

b DC and factory

Factory 26 Street 8, Vietnam –

Singapore Industrial Zone,Binh Duong Province,Vietnam

1500

DC 54 Le Duan Street, Ninh

Table 1: DCs and factory location

All of these statistics above are real, except for the capacity of the factory From ourresearch, the area of the factory is up to 60000 m³, with the warehouse can be about

20000 m³ [3] This factory is in charge of producing 10 product lines of Kinh Do [4] ,which leads us to the assumption that the capacity of the factory for 3 mentionedproducts can be 1500 m³

c Products

There are 3 products focused on in our report The selling price comes fromunpublicized costs for retailers revealed by Mondelez Kinh Do Company Withoutreal data for cost, we assume that the total cost takes up about 65% selling price

d Demand

We collect real data on the demand of 100 retailers in Ninh Thuan Province for our 3target products All the demands are assumed to start on the first day of theinvestigated period, 01/01/2022 For the sake of convenience, we calculate thedemand of each site, which is unique, irregular and based on various complexmechanisms After that, we convert these data into only one pattern - periodic period,which means orders are placed periodically Followingly, order interval and quantityper order are customized based on reality so that the total demand of each customerimported into Anylogistic is equivalent to the real demand Despite this simplification,

we assure that there is no significant difference between real statistics and in-usestatistics For 100 customers and 3 products, there are 300 demands correspondingly

For further detail, read Appendix 2

interval (days)

Quantity (boxes)

Expected Lead Time (days)

Table 4: An example for Demand

e Facility expenses

Facility expenses are assumed from the fact that expenses for renting a factory are110.000 VND/ m2 [5] Empirically, each box of products contains an average of 60packs and occupies 0.05 m³ All this evidence leads us to the assumption that eachbox of products takes up about 500 VND for carrying cost per box per day and weuse this statistic for carrying cost in DC also

Facility Expense Type Value Currency Time Product

Unit Unyou it

Table 5: Facility Expenses

f Inventory

Because of the lack of information accessibility, we assume the inventory policy For

DC, we apply Inventory policy Min Max with Safety stock Safety stock can becalculated by the formula: [6]

Safety stock = (Maximum amount of sales * Maximum lead time) –

(Average amount of sales * Average lead time)

Although all of these component data should be calculated daily, we cannot reachexactly this requirement So, we use the accumulated data from 01/01/2022 to31/12/2022 (which is empirically obtained) to calculate and yield the number 157boxes A periodic check is assumed according to the typical expiry day of packagedfresh bread (which is mentioned before) is 7 to 10 days

Table 6: Inventory

g Vehicles

Capacity of truck and van in box unit derive from articles shown in References [7]

Name Capacity (Boxes) Speed

[9] For simplification, we use 27000 VND/litre as fuel cost

Fuel consumption rate: A typical truck with a truckload of 2 tons has a fuelconsumption rate of 9 litres per 100 kilometres

- So, fuel cost per distance unit is:

- Also, we found that on the path from the factory to DC, there is one BOTstation and it charges 59000VND [10] for the truckload mentioned above

- For labour cost, which is the wage for drivers scientifically, is assumed to be

320000 VND

So, the total fixed cost is calculated and rounded to be 1130000 VND

ID From To Cost Calculation Parameters

(VND)

Vehicle Type

Transportation time (days)

Path 1 DC

Location

Customer group

Distance-based with cost per stop

2000*distance + 30000*stop

Path 2 Factory

Location

DC Location

Fixed delivery 1130000 Truck 1

Table 8: Paths

i Processing cost

"Processing cost" refers to the expenses incurred during the production ormanufacturing process of goods or the rendering of services It includes all the costsassociated with converting raw materials into finished products or delivering services

to customers These costs can encompass a wide range of expenses, such as labourcosts, raw material costs, energy costs, equipment maintenance, and overhead costsrelated to the production facility

(VND)

Factory BURGER BÒ Inbound Shipment Processing Boxes 3000

Processing

Boxes 3000

Factory PIZZA XÚC XÍCH Inbound Shipment Processing Boxes 3000

DC PIZZA XÚC XÍCH Outbound Shipment

Factory (All

products)

Inbound Shipment Processing

Shipment 3 – 5

Factory (All

products)

Outbound Shipment Processing

Shipment 3 – 5

products)

Inbound Shipment Processing

Shipment 2 – 3

products)

Outbound Shipment Processing

Parameter Priority

DC Customer

group

(Allproducts)

Sourcing 2 Customer group (All

products)

Closest (DynamicSources)

Table 13: Sourcing

2.2.2 Process and output

In the operation process, we use Simulation as the only tool for our purposerepresented above

“Simulation experiment is not an analytical method of optimizing your supply chain(as compared to GFA and Network Optimization experiments), time is relevant in thiscase The experiment is used to properly configure inventory policies, find optimalproduct stock volume, eliminate the possibility of lost orders and execute what-ifscenarios to see how the changes you make affect the outcome.” [11]

We import statistics collected or hypothesized as presented above into Anylogistix as

a Simulation scenario, to visualise our base case, or the existing supply chain Hereare some settings before running the experiment:

Figure 5: Simulation experiment setting

Here are some results:

Figure 6: Base case - simulation result (1)

Figure 7: Base case - simulation result (2)

Figure 8: Base case - simulation result (3)

Figure 9: Base case - simulation result - profit and lost

Among the statistics shown above, we delve into statistics about profit and revenue inthe graph below:

Figure 10: Base case - Revenue and Profit graph

With the profit rate of about 41%, Chapter 3 will demonstrate several scenarios whichare supposed to increase this rate

2.2.3 Validation

According to the statistics of Kinh Do Company, the aggregate revenue from 5 mainproduct lines (Solite, Kinh Do Biscuit, Fresh Kinh Do, Kinh Do Bread and Kinh DoCake) in Ninh Thuan Province in 2022 is 12.423.640.000 VND Although the exactreal statistic of revenue of three targeted products is inaccessible, we can compare therevenue generated by Anylogistix for three products to the real total revenue Roughcalculation gives the revenue of the three mentioned products account for about 25%

of the total revenue This significant proportion validates the popularity of “Sandwichchà bông”, “Burger bò” and “Pizza xúc xích” as compared to other products of Kinh

Do For further validation tests, real revenue and real cost of three products arerequired

CHAP 3: EXPERIMENTAL DESIGN

3.1 Scenario 1: Add one more DC

Figure 11: Pre-existing supply chain

Considering the real supply chain, it is transparent that the demands of 100 customersare in charge by just 1 DC, which may lead to much pressure in terms of lead time andinventory for the DC We conclude that the supply chain can be optimized byadjusting the structure of the supply chain We decide to put a hypothesis that there isone more DC in Ninh Thuan Province

3.1.1 Process and output

To find the most ideal location for the additional DC, we use Greenfield Analysis(GFA)

The GFA, also known as the focal point analysis, is a common method fordetermining optimal locations for new plants[12] The issues we need to consider in aGreenfield analysis are our customers' locations, the distances from our warehouses toour customers, and what our customers require of our products GFA is used to findthe optimal location within a network for setting up a new production facility orwarehouse, while Brown Field analysis using the same technique can be used to adaptexisting networks To determine the optimal location for a manufacturing or storagefacility, the point is identified where the sum of the distances from all suppliers to thefactory (demand point), weighted by the volume of product flow between eachsupplier and the potential factory, is minimal To determine the optimal warehouse

location, the distances from the customer to the warehouse are also calculated,weighted with their respective requirements.

In the GFA section, we entered the data which are the same as the base case, including

Customers, DCs and Factories, Demand, Location, Periods, Products, Sourcing and Units

After that, we started to run the GFA experiment

3.1.1.1 GFA experiment

- Set the “Experiment duration” for “all periods”, which refers to the default

time period (01/01/2022 - 31/12/2022)

Figure 12: Scenario 1 - GFA setting - period

- Let the experiment be analyzed by “Number of sites” and set 1 site,

specifically It means that one additional DC with an optimized location willappear

Figure 13: Scenario 1 - GFA setting - number of sites

- Set the “suppliers to sites transportation” discount, %” and “distance step for stats” to be the same as the default setting.

Figure 14: Scenario 1 - GFA setting - default constraints

- Finally, we change the “products stats unit” into “boxes” and the “distance stats unit” into “kilometres”.

Figure 15: Scenario 1 - GFA setting - unit setting

● Result:

After adjusting all the GFA settings, we start to run the experiment The experimentthen gives us a new DC which is located at the latitude ( 11.775 ) and the longitude( 108.79 ) This new DC is responsible for the delivery of 16 customers

Figure 16: Scenario 1 -GFA result - New DC location (1)

Figure 17: Scenario 1 - GFA result - New DC location (2)

Figure 18: Scenario 1 - GFA result - New DC location (3)

For the next phase, we convert this result into Simulation as a new scenario, toinvestigate some categories we are looking for

3.1.1.2 Simulation experiment

- Firstly, from an overall view, the new DC (GFA DC) is quite near the initialone (DC), so we configure the network so that products will originate from thefactory to the DC At DC, products are transported toward 2 paths: one pathtoward the customer group of DC and another path toward the GFA DC, andthen products are transported to the GFA DC customer group

Figure 19: Scenario 1 - Simulation experiment - structure

- Thanks to GFA, we have two automatically-generated customer groups for

each DC, which are shown in Table Group

Figure 20: Scenario 1 - Simulation experiment - group

- At the DCs and Factory section, we turn on the GFA DC and include the GFA

DC in our chain We then add the capacity for this new DC, which is supposed

to be 300 m³, because it just serves only 16 over 100 customer sites

Figure 21: Scenario 1 - Simulation experiment - DCs and Factory

- In the Facility Expenses section, we add the carrying cost of the GFA DC like

the old DC, which is 500 VND for 1 box in 1 day and we also add facility cost,which is 200 VND for 1 box in 1 day

Figure 22: Scenario 1 - Simulation experiment - Facility expenses

- At the Inventory section, we apply the min-max policy with safety stock

policy to all the 3 products in the GFA DC in 7 days In more detail:

+ PIZZA XÚC XÍCH: Min: 30 boxes, Max: 70 boxes, Safety stock: 20and The initial stock: 50 boxes

+ SANDWICH CHÀ BÔNG: Min: 25 boxes, Max: 60 boxes, Safetystock: 15 and The initial stock: 40 boxes

+ BURGER BÒ: Min: 25 boxes, Max: 50 boxes, Safety stock: 15 and Theinitial stock: 40 boxes

Figure 23: Scenario 1 - Simulation experiment - Inventory

- In the Paths section, we add a new path from DC to GFA DC with fixed

delivery calculation at 300000 VND and a new path from GFA DC to GFA DCCustomer group with distance-based with cost per stop calculation at 2000VND for 1 km plus 30000 VND per stop

From To Cost

Calculation

Cost CalculationParameters

Distance-basedwith cost perstop

2000*distance +30000*cost

Distance-basedwith cost perstop

2000*distance +30000*cost

Van

Figure 24: Scenario 1 - Simulation experiment - Paths

- At the Processing Cost section, we add all 3 products for GFA DC with

outbound shipment at 3000 VND per box

Figure 25: Scenario 1 - Simulation experiment - Processing cost

- At the Processing Time section, we add outbound shipment processing with a

shipment unit and the time is uniform from 2 to 3 hours

Figure 26: Scenario 1 - Simulation experiment - Processing time

- At the Shipping section, from GFA DC to GFA DC Customer group, the

vehicle type is a van with Less than Truckload type

Figure 27: Scenario 1 - Simulation experiment - Shipping

- In the Sourcing section, we add a sourcing path from DC to GFA DC and from

GFA DC to GFA DC Customer group with the closest fixed source for all 3types of products

Figure 28: Scenario 1 - Simulation experiment - Sourcing

After adjusting data as shown above, we run a simulation experiment and gain the final result

3.1.2 Result of Scenario 1

After fixing and changing all the information above, we begin the simulationexperiment and receive the results for the dashboard and profit and loss

- The Dashboard:

Figure 29: Scenario 1 - Simulation result - Dashboard

- The Profit and Loss:

Figure 30: Scenario 1 - Simulation result - Profit and lost

Figure 31: Scenario 1 - Simulation result - Revenue and Profit graph

The revenue of this system reaches 3,349,381,000 VND Then, after spending forfacility cost, inbound processing, inventory carrying, outbound processing, productioncost and transportation cost, the total profit is 1,329,322,254.479 which accounts forabout 39.68% of revenue

Over the first half of the year, the profit is almost negative It is perhaps because thesuppliers are excluded from our system, so the factory has to order raw materialsinitially, leading to many backlogs