Frontiers in Computing and Intelligent Systems 
ISSN: 2832-6024 | Vol. 1, No. 1, 2022 

 

66 

Design and Application of Logistics Robot based on RPA 

Zehua Zhang, Ziheng Huang 

Chongqing University of Posts and Telecommunications 

 

Abstract: In the logistics management process, tedious and repetitive tasks are very common, which is inseparable from a 

large number of manual data entry and processing of a large number of orders to complete the daily logistics workflow. The work 

of entering data and processing orders takes a lot of time and high repetition. The application of RPA robot (Robotic Process 

Automation) is particularly important, which can reduce a lot of labor cost and time cost, improve accuracy, and is of great 

significance for logistics management. This paper designs a RPA robot framework for logistics management, which is directly 

used in the development tool Uibot Creator. Developers carry out the second development on the basis of this framework, which 

can simplify the development process, improve the efficiency of development, save project preparation time, and improve the 

efficiency of logistics operation and management. The core functions of the framework include logistics web portal framework, 

logistics information capture, input, logistics documents, mail automation and other frameworks. Finally, the designed Robotic 

Process Automation framework is applied to the logistics management process to verify the effect of the optimization.  

Keywords: Robotic Process Automation, Logistics management, Development framework design, Application. 

 

1. Research background 

The analysis of the World Economic Forum shows that 

with the digital transformation of the transportation industry, 

the value of logistics enterprises will exceed 10 trillion RMB, 

and the social benefits will reach 16 trillion RMB. In the 

process of digital transformation of logistics enterprises, how 

to simplify the logistics management process and speed up the 

core business, how to reduce costs and further improve 

service quality will determine the future of logistics 

enterprises. At this time, RPA (Robotic Process Automation) 

technology has the opportunity to exert its talents. RPA 

software tools simulate the processes repeated by logistics 

managers every day to improve the efficiency of logistics 

operation and management, such as simulating the process of 

simulating operators to enter a large number of data manually 

and deal with a large number of orders. 

The business process digital transformation of logistics 

enterprises is in a period of rapid development, at this time 

the basis of the transformation and development comes from 

various factors of the technical dimension, and the Robotic 

Process Automation technology will be an important part of 

the digital transformation of logistics operation and 

management. Cheng Ping etc. [2-8] used robotic process 

automation technology to optimize the financial management 

process, such as fixed assets management process, accounts 

receivable management process, tax management process and 

so on. RPA robot was designed for the process to improve the 

management efficiency. Tian Gaoliang studied the application 

of financial robot based on RPA technology. Quan Wenju 

applies RPA financial robot to electric power enterprises, and 

explores the significance of RPA in improving the 

management level of electric power enterprises. He Hong etc 

applied RPA to insurance companies, which reduced the 

repetitive work of financial managers. Liu Qingying  

analyzed the influence of RPA robot on the financial 

management of enterprises. Lu Ying  applies intelligent 

finance to financial management in the field of real 

estate.Jiang Mingyu  designed a template framework for 

developers based on RPA technology, and completed the 

automatic generation of An enterprise accounting reports 

based on the framework, which greatly improved the 

development efficiency of the process. 

Yu Yanling analyzed the influence of digital transformation 

on enterprise logistics marketing mode, and Yu Lei put 

forward the intelligent management mode of logistics and 

commodities with intelligent LOT network. Zhang Daochen 

analyzed the problems existing in the operation of logistics 

parks in China. The research of the above scholars shows the 

importance of the digital transformation of logistics 

enterprises, and the logistics management process needs to be 

further optimized. Robotic Process Automation technology 

will solve a large number of problems in logistics 

management process and improve the efficiency of logistics 

management. 

2. Business pain points in the process of 
logistics management 

Logistics operation and management process is inseparable 

from a large number of manual data entry and processing of a 

large number of orders to complete the daily logistics 

workflow, data entry and order processing work takes a lot of 

time and high repetition. It mainly involves inventory 

management, order management, logistics information 

management, transportation management and other 

businesses. at present, these services require a large number 

of repeated manual operations, which are easy to cause data 

errors and untimely updates. at the same time, there are 

complex customer relationships in logistics enterprise 

management that need to be dealt with. 

2.1. Inventory management 

2.1.1. Purchase and storage process 

For organizations with multiple suppliers and engaged in a 

large number of production activities, it is very important to 

purchase all kinds of materials, and hundreds of activities may 

occur every day. Before the arrival of the wave of intelligence 

in the industrial manufacturing industry, corporate 

organizations could only manually update the materials they 

needed every day and submit orders to suppliers. This kind of 

operation is not only time-consuming and laborious, tedious 



 

67 

and tiring, but also destroys the enthusiasm of employees 

because of the boring work. 

2.1.2. Inventory management process 

Before the RPA robot participates, the staff responsible for 

inventory and order processing need to log in to the business 

system, check the inventory levels of different warehouses 

one by one, and deliver the inventory list after screening the 

goods. Then, after updating the warehouse, the staff need to 

fill out the invoice one by one, confirm the inspection one by 

one, and finally send the order to the logistics supplier system. 

Manual execution cannot monitor inventory in real time, find 

missing materials and send orders in time, which will lead to 

untimely delivery of orders, and then affect the industrial 

supply chain and increase the risk of sales loss. 

2.2. Order management 

With the increasing number of orders, the status of logistics 

orders needs to be updated in real time every day. Operators 

need to sort out the unprocessed order numbers in advance, 

summarize them into the Excel table, then manually log in to 

the relevant logistics system, and copy these order numbers 

to the corresponding area of the system page. Operators 

manually handle hundreds of order numbers every day, which 

is a huge workload. 

For booking documents, the field is more than 50, and it 

takes more than five minutes for the operator to enter an order, 

and it needs to be checked repeatedly. Coupled with the 

different sources of booking manifests, the format cannot be 

unified, which makes it difficult to import in bulk. The 

workload is so heavy that it takes more time to complete the 

booking data entry process during the peak season, and 

employees have to work overtime. 

2.3. Delivery and sorting 

Party A's supply chain management personnel need to 

manually query the relevant information in the ERP system 

according to the established conditions, grab the goods list 

information that has been stored and ready for transportation, 

and generate the waybill, which contains the recipient, 

address, mode of transportation and other information, and 

finally send the order to the logistics supplier's OMS or TMS 

system. This process has a lot of manual operations, a large 

amount of data, high repeatability and error-prone. 

2.4. Logistics information management 

2.4.1. Logistics staff 

Logistics operators need to log in to the logistics 

information query system every day, follow up and monitor 

the status according to the logistics status information, 

download and input the data at the end of the signed status, 

and follow up and continue to track the unsigned status. for 

the need to manually check and track the abnormal logistics 

information, you need to download the status category 

manually, in addition, you need to enter the information 

completion system after the data download. Manual review is 

a waste of time. 

Logistics automatic return personnel need to log in to the 

logistics return system every day, upload the finished data 

through the excel form, manually check the abnormal form 

information, and do a good job of data collation for normal 

information, but because the abnormal form does not report 

an error reminder, it needs to be checked manually, and the 

cost of manual review is high. 

Logistics customs declaration personnel need to log in to 

the to-do and done task list of the customs declaration 

management platform system every day, submit and view the 

logistics declaration information, match and verify the 

information of multiple approval stages, download the 

information that meets the customs declaration requirements 

through the background system, and open the Excel form, fill 

in the relevant information such as the return receipt and 

processing time of the customs declaration and audit, and 

calculate the time spent in all stages. Finally, it is checked and 

checked manually by sending e-mail. 

2.4.2. Logistics management personnel 

After sending the goods, the supply chain management 

personnel of Party A need to manually log in to the system or 

website provided by the logistics supplier and obtain the 

logistics status information from the logistics supplier 

through the logistics order number. Update the information to 

various logistics information platforms, so that customers can 

track the current logistics information in a timely manner. In 

this process, the workload of logistics management personnel 

is huge and repetitive. 

2.5. Transport management 

In the process of transportation, motor vehicle violations 

are inevitable, which requires logistics and warehousing 

companies to query and deal with the relevant work. There 

are a large number of repetitive operations in this process, 

which require a lot of labor costs and time costs, and the 

operation is not timely enough, it will produce a large number 

of late fees, resulting in greater losses. When the transport 

goods are dangerous goods, the relevant management 

personnel need to fill in a large number of application forms 

before the transportation of dangerous goods, and need to log 

into the road transport safety supervision system constantly to 

apply, the application process is repetitive and tedious, large 

and error-prone. 

2.6. Financial problems in logistics 

management 

Whether we are sending express logistics or receiving 

express logistics, we will often face the need to issue invoices, 

while we have to send express delivery and deal with invoices 

the whole thing is very messy. 

2.7. Logistics bidding 

The managers of logistics suppliers need to log in regularly 

to the bidding system or website provided by Party A to 

further calculate the bidding information of the project, and 

then manually upload the relevant bidding documents and 

manually calculate the quotation. This workflow has a large 

workload, high repeatability, is prone to errors, and requires a 

short time to complete, so it is difficult to submit relevant 

information in time. 

3. The design of logistics management 
robot based on RPA 

The above problems in the logistics management process 

can be solved and optimized through RPA (Robotic Process 

Automation). RPA refers to the use of software with AI 

(artificial intelligence) and ML (machine learning) functions 

to deal with some repetitive and heavy workload tasks that 

need to be performed manually. In this paper, a RPA robot 

framework for logistics management is designed to solve the 

above problems. Developers carry out the second 



 

68 

development on the basis of this framework, which can 

simplify the development process, improve the efficiency of 

development, save project preparation time, and improve the 

efficiency of logistics operation and management. 

This article will use Uibot Creator, a common development 

tool of RPA in China. The core module of the framework 

includes execution process module, initialization module, 

data processing module and exception handling module. The 

execution process block contains various core functions of the 

logistics management process, and each module has a 

reserved interface to facilitate developers to use the 

framework directly. The development main framework is 

shown in figure 1: 

 

 

Figure .1 Development Agent Framework 

In this framework, initialization, target program 

initialization, first run judgment and retry judgment are 

configured as initialization function modules, exception 

handling and retry counting process blocks are exception 

handling function modules, obtaining new data and judging 

whether there is a data flow block as data processing function 

module, execution process block as sub-process, which 

contains the main functions needed by logistics management 

process.Contains logistics web portal framework, logistics 

information capture, input, logistics documents, mail 

automation and other frameworks, the specific process is as 

follows: 

3.1. Initialization module 

The initialization module mainly includes configuration 

initialization and object program initialization. 

3.1.1. Configuration initialization 

 
Variable Name Value 

g_bRetryStatus False 

g_dicTransactionItem Null 

g_dicConfigData {} 

g_iCount 0 

g_iRetryNumber 0 

g_bFirstRun True 

Figure .2 Initial values of global variables 

The configuration initialization process block is used to 

initialize the parameters, set the log level, set and read the 

configuration file. The parameters are mainly global variables 

g_iCount (initial value of retry count), g_iRetryNumber 

(number of retries), g_bRetryStatus (initial value of 

transaction item retry state), g_dicTransactionItem 

(transaction item data), g_bFirstRun (identification of the first 

run). The initial value is shown in figure 2. The relevant data 

of the logistics management process will be put into the 

g_dicTransactionItem and configuration files, such as 

inventory information, order information, logistics status, 

financial information, logistics indicators and so on. 

3.1.2. Object program initialization 

The initialization of the target program is to initialize the 

environment before the process runs, read the related fields 

from the flow chart variable g_dicConfigData, deal with them, 

check whether the target program of the RPA process has been 

installed and other necessary preconditions, judge whether the 

robot environment meets the requirements, close the browser, 

close the logistics system, close the software, and so on.After 

the initialization of the target program, it will determine 

whether to run it for the first time, and if it runs for the first 

time, it will carry out data processing, otherwise it will choose 

to run the main program directly or get new data according to 

the limit of the number of retries. 

3.2. Exception handling module 

When there is an exception in the process block such as 

configuration initialization, target program initialization and 

data processing, the exception handling process block will be 

carried out to complete the process of writing error log, 

sending e-mail to notify relevant personnel, screenshot and 

other processes; when there is an exception in the execution 

process block, it will enter the retry counting process block to 

complete the data update, screenshot, timing and other 

processes of g_bRetryStatus and other variables.The main 

code is as follows: 

Log.Error("an exception occurred in the process, 

processing!") 

Dim SystemException = Null 

SystemException = $BlockInput //system variabl 

Log.Error(SystemException) 

3.3. Data processing module 

Get the data and assign it to the flowchart variable 

g_dicTransactionItem for global use. The subflow needs to be 

used after input. The data can be obtained by consuming a 

data queue message specified in Uibot Commander, reading 

a specific row in the specified Excel table, reading the 

specified field data in the specified database, reading an email 

in the specified mailbox, and obtaining the returned data 

through webAPI.Or choose other methods of data acquisition 

according to the needs of logistics managers.It must contain 

the following code: 

g_bFirstRun = False //enter this process block, and the 

initial running status needs to be updated to False. This 

command cannot be deleted. 

g_dicTransactionItem = Null //initialization is required 

before getting the data of a new transaction. This command 

cannot be deleted. 

3.4. Logistics web Portal Framework 

This framework is mainly aimed at the acquisition and 

input of logistics information in the logistics web portal, using 

RPA and python plug-ins to complete this process.This 

process automates the browser with the selenium feature of 

python, and the following python packages need to be 

imported: 



 

69 

from selenium import webdriver 

from selenium.webdriver.support.ui import 

WebDriverWait 

from selenium.webdriver.support import 

expected_conditions as EC 

from selenium.webdriver.common.by import By 

from selenium.webdriver.chrome.options import Options 

After importing the python package, you can automatically 

log in to the browser and go to the designated logistics portal 

for related operations. You can use a headless browser to 

automatically run the process in the background, and the 

browser will not be displayed in the windows page. This 

cannot be done by RPA. The core code of python is as follows: 

def python_selenium(content): 

    #headless browser configuration 

    chrome_options = Options () 

    chrome_options.add_argument('--headless') 

    chrome_options.add_argument('--disable-gpu') 

    # instantiate a browser object (the driver passed into the 

browser) 

    bro = 

webdriver.Chrome(executable_path='./chromedriver', 

options=chrome_options) 

# Let the browser find a request corresponding to the 

specified url 

    bro.get('https://*****.com/') 

   # locate the relevant buttons in the browser 

ele = WebDriverWait(bro, 

60).until(EC.presence_of_element_located(*****)) 

# enter the logistics information to be queried in the input 

box 

    ele.send_keys(content) 

# wait a while for the relevant content to appear 

    sleep (2) 

# obtain logistics information 

new = bro. find_element_by_class_name('textpanel-target-

textblock').text 

# close the browser 

    bro.quit() 

    return new 

Import the python_selenium function into Uibot through 

the above Python plug-in. When you do not need to run the 

browser in the background, you can also directly use the RPA 

framework to operate the logistics web portal website. The 

specific process is shown in figure 3: 

 

 
Figure .3 Logistics web portal flow 

3.4.1. Log in to the logistics Web portal process 

This process is divided into three steps: logging in to the 

website, closing the pop-up window, logging in to the account, 

logging in the URL address of the website in the login process, 

and maximizing the browser window. There is an interface at 

the properties of the browser window to facilitate developers 

to perform secondary development according to the actual 

needs to complete the corresponding operations. The core 

code of the login process is as follows: 

Dim hWeb = "" 

hWeb=WebBrowser.Create("chrome","about: 

blank",30000, {"bContinueOnError”: 

False,"iDelayAfter":300,"iDelayBefore":200,"sBrowserPath

":"","sStartArgs":""}) 

Window.Show({"wnd":[{"cls":"Chrome_WidgetWin_1","

app":"chrome"}]},"max") 

The process of closing the pop-up window has a relevant 

interface set at the pop-up window properties. Developers can 

configure relevant parameters according to the specific 

conditions of the website and set the closure of multiple 

windows to prevent the pop-up window interface from 

affecting the next relevant operations. the login process 

mainly includes judging whether you have logged in to the 

logistics administrator account and automatically clicking the 

login button to enter the account login interface if not logged 

in.Enter the specified account and password (g_dictGlobal 

["usName"] and g_dictGlobal ["usPassword"]) set in the 

initialization process in the account password input box, and 

click Log in to complete the account login operation. This 

process has relevant interfaces in the account password input 

box, login button and other locations, and the developer 

configures it according to the interface elements of the input 

box of the login interface. 

3.4.2. Logistics data capture process 

This process is carried out on the basis of opening the 

logistics website and logging in, and grabs the relevant 

logistics data according to the demand, including in and out 

information, logistics order information, logistics status 

information and so on. The capture time can also be run 

regularly according to the needs of logistics managers, and 

the time interval can also be set to ensure timely update of 

logistics information. The main command used is data capture. 

As shown in figure 4, after clicking the data crawling 

command, the page shown in figure 5 will appear. Click to 

select the target and select two targets at the same level, and 

you can crawl the data of the relevant tables and web pages. 

But this method is only applicable to the data with tables or 

rules, so this process framework also provides other relevant 

commands to meet the needs of crawling data. These 

commands require developers to do more secondary 

development according to the characteristics of the web page. 

The main commands that need to be used are getting the 

element text, getting the child elements, traversing the array 

and so on. There is an interface in the element location, and 

the final data will be stored in the variable arrRet. 

 

 
Figure .4 Location of data crawling command 

 

 
Figure .5 Data crawl command page 

3.4.3. Logistics data entry process 

This process is carried out on the basis of opening the 

logistics website and logging in, inputting relevant logistics 

data according to demand, including in and out of warehouse 

information, logistics order information, logistics status 

information and so on. The entry time can also be entered 

regularly according to the needs of logistics managers, and 

the entry interval can also be set to ensure the timely updating 

of logistics information. The main commands used are input 

text in the target, setting element check, setting element 

selection, obtaining child elements, traversing sub-elements 



 

70 

input in turn, etc., and there is an interface in the element 

location. 

3.5. Logistics information capture framework 

This framework aims at grasping logistics information in 

relevant documents, systems and databases, and grabbing 

relevant logistics data according to demand, including 

logistics system background information, in-and-out 

information, logistics order information, logistics status 

information and other data. the operation of grabbing 

information can be grabbed again when the logistics 

information is updated to ensure the timely update of the 

logistics information. This framework includes opening files, 

systems, logistics information capture, logistics information 

processing and other processes, as shown in figure 6: 

 

 
Figure .6 Logistics information capture process 

The process of opening files and systems is mainly to 

prepare for the process of grabbing logistics information, 

opening Excel files, databases and systems containing 

logistics information. The main commands in this process 

include selecting files dialog box, obtaining file list, opening 

files in turn, opening corresponding system software, and so 

on. 

At the end of the process of opening the file and the system, 

the logistics information in the file or system will be grabbed, 

and the operation of reading cells, rows, columns and areas 

will be carried out for the Excel table, and the same operation 

as the browser will be carried out for the system, obtaining 

the target sub-elements, traversing the sub-element array, 

obtaining the target element text, etc., for the database, the 

SQL statement can be executed directly to fetch the logistics 

information of the database. The framework has interfaces in 

the cell, row, column, area and element location of the Excel, 

which requires developers to choose the corresponding 

location according to the requirements. 

After completing the logistics data capture process, the 

captured data needs to be processed and stored. This 

framework stores the captured information in Excel, and 

developers can change it according to their actual needs. The 

main commands involved in the data processing process are 

building data table, data slicing, data deduplication, data 

sorting, data table merging, array processing, etc., as shown 

in figure 7: 

 

 
Figure .7 Data processing commands 

After completing the data processing, the processed 

logistics data is stored in the Excel table. The main commands 

involved are creating Excel table, writing regions, cells, rows 

and columns, creating worksheets, saving and closing Excel, 

and so on. Developers can deal with Excel tables according to 

their actual needs. 

3.6. Logistics information input framework 

This framework mainly includes two processes: opening 

the Excel file or system that needs to be entered and inputting 

information, and inputting relevant logistics information 

according to demand, including data such as in-and-out 

information, logistics order information, logistics status 

information, etc., the entry time can also be entered regularly 

according to the needs of logistics managers, and the input 

interval can also be set to ensure timely updating of logistics 

information. The main commands used are operated with the 

browser, including entering text in the target, setting element 

check, setting element selection, obtaining sub-elements, 

entering sub-elements in turn, etc., and setting an interface in 

the element location. 

3.7. Mail Automation Framework 

This framework mainly includes real-time monitoring, 

completion notification, error notification and other processes. 

When the whole RPA process is running, email is sent after 

the corresponding operation is completed, and logistics 

managers are notified. After the relevant logistics information 

is updated, the relevant personnel can also be informed by e-

mail. When the whole process ends, or when the process 

reports an error, the operator of the process can also be 

informed by email. The main commands involved in this 

framework are getting the current time, getting the element 

text, judging the text change, sending mail, trying to execute, 

sending error mail and so on. 

3.8. Logistics document framework 

This framework mainly processes some documents 

through the Mage AI function of UiBot. The document 

recognition types include text recognition, form recognition, 

general multi-bill recognition (commonly used), card 

recognition and so on. Before the start of the process, select 

the Mage AI function shown in figure 8. Then there will be a 

pop-up window shown in figure 9. after selecting the 

corresponding recognition type according to the type of 

logistics document, enter the selection of image source. And 

according to the actual needs to select the type of documents 

for further processing. 

 
Figure .8 Mage AI 

 
Figure .9 Mage AI configuration page 

4. The application of RPA robot in 
logistics management process. 

Based on the above RPA robot framework, the related 

processes of logistics management are optimized and applied 

to inventory management, order management, shipping and 

sorting, logistics information management, transportation 

management, financial problems in logistics management, 

logistics bidding and so on. The framework mentioned above 

can be used in these processes, including initialization module, 

exception handling module and data processing module. Then 

use other related functional frameworks according to the 



 

71 

requirements. 

4.1. Automation of inventory management 

4.1.1. Purchase and storage process 

In this process, the RPA robot can monitor the inventory 

level in real time and notify the team before the inventory 

level becomes low, so that the team can order products in 

advance and generate real-time reports to optimize inventory 

demand.After participating in the inventory management 

robot, it can automatically log in to the business system, 

monitor the inventory level and generate a report, and then 

grab the delivery list.RPA robots can synchronize information 

with systems such as the company's CRM.Based on the 

generated report information, the inventory management 

robot can send mail to communicate with the supplier to place 

an order.The purchasing manager approves the request 

according to the email notification, and the RPA robot can 

automatically send the order to the logistics supplier system. 

This process needs to use the framework of logistics web 

portal framework, logistics information capture framework, 

mail automation framework.First of all, you need to use the 

login logistics portal website or system process in the logistics 

web portal framework, the user needs to add the URL address 

of the website or the location of the logistics system in the 

configuration file, and the secondary developer needs to 

configure elements such as the input box of the login interface 

at the interface of the framework, as shown in figure 10.You 

can directly select the login button where you click the target, 

or configure the element of html in the properties. 

 

 
Figure .10 Login to the logistics system configuration page 

After automatically logging in to the website or system, 

automatically click to enter the inventory page, as shown in 

figure 11, to grab the inventory information. The information 

capture process needs to use the framework of logistics 

information capture, and directly use the data capture function, 

such as the function page shown in figure 12, to grab the 

inventory in the table, according to the actual situation of 

logistics enterprises, grab inventory information regularly, 

and store the data in the inventory level report after capture. 

 
Figure .11 Inventory Management Page 

 

 
Figure .12 Data capture page 

Based on the report information of inventory level, the mail 

automation framework is used to notify suppliers and 

purchasing managers to place orders, in which developers are 

required to configure mailbox SMTP server, port, account 

number, password, sender, recipient, title, body, attachment, 

etc. as shown in figure 13, the recipient can be the relevant 

personnel of the supplier or the purchasing manager, and the 

generated inventory level report can be added to the email 

attachment. 

The RPA robot based on the framework designed in this 

paper can replace the manual way to update the materials 

needed every day in the purchase and storage process. The 

RPA robot can update the inventory online all day, 24 hours a 

day, generate a report to remind the stock when it is out of 

stock or higher than the upper limit, and automatically fill in 

the invoice.In this way, the enterprise's inventory 

management can be in the best state at all times.RPA robot 

can effectively reduce the risk of sales loss caused by 

inventory or raw material shortage, ensure that enterprises 

maintain the best inventory level, and effectively manage the 

supply of all kinds of products in times of peak and low 

demand. 

 

 
Figure .13 Mailbox configuration page 

4.1.2. Inventory management process 

In this process, the RPA robot can automatically log in to 

the business system, check the inventory levels of different 

warehouses one by one, and automatically screen the goods 

and deliver the inventory list. After updating the warehouse, 

the RPA robot can automatically fill out the invoice, confirm 

the inspection one by one, and finally automatically send the 

order to the logistics supplier system. 

The frameworks that need to be used in this process are 

logistics web portal framework, logistics information capture 

framework, logistics information input framework and mail 

automation framework. The development step of this process 

is roughly similar to the purchasing process, which is also 

automatically logged into the system to capture inventory data. 

The difference is that after capturing inventory data, you need 

to use the framework of logistics information input. Fill the 

relevant logistics information into the system shown in figure 

14 and confirm it. The secondary developer needs to 



 

72 

configure the attributes of the system input location in the 

location shown in figure 15. You can select the input target at 

the search target that gets the location of the child element, or 

configure it in the target location in the right attribute. Finally, 

the process uses the mail automation framework to 

automatically send mail to the vendor. 

 

 
Figure .14 Storage system page 

 

 
Figure .15 Input information configuration 

The secondary development of RPA robot based on the 

framework designed in this paper can solve the problem that 

manual execution cannot monitor inventory in real time. RPA 

robot can find the missing materials and send orders in time, 

provide product-based lifecycle management services, 

provide partners with order management and other services, 

and form a scene ecological data closed loop. To achieve 

process standardization, reduce the occurrence of verification 

matching errors and other problems, through the digital staff 

to greatly improve the efficiency of verification, through the 

docking of internal and external systems, without the need for 

integration cost and time. Through the price collection and 

feedback in different periods, it can provide a strong basis for 

future production, design and sales. 

4.2. Automation of order management 

4.2.1. Waybill automation 

We can use the RPA robot to handle a large number of 

online couriers or pickups, so that we can better publish this 

information to couriers in each area. Then, according to the 

data in the company database, automatically enter customer 

information, process payment, send e-mail to confirm and 

place orders, automate the process, and deal with abnormal 

parts in batches. The RPA robot automatically sends the 

information of the goods in the order system every day, and 

then obtains all kinds of information of the goods (such as 

parameters, transportation requirements, consignee 

information, etc.) from different systems, and finally uses the 

specified format to automatically make the logistics delivery 

list. 

With the increasing number of orders, the status of logistics 

orders needs to be updated in real time every day. The RPA 

robot can monitor the status of the waybill in real time, and 

automatically update the relevant order transfer details in the 

system, automatically summarize them into the Excel table, 

and automatically transfer them to the logistics system. The 

RPA robot can work with intelligent routing systems based on 

artificial intelligence, coordinate multiple logistics partners, 

such as road transportation, shipping and air transportation, 

and automatically assign delivery companies and methods 

according to the specific place of delivery of the product. The 

RPA robot can send email to customers in time when there is 

a delay in logistics, and enhance the customer experience 

through real-time updating of orders. 

This process mainly uses the logistics information capture 

framework, mail automation framework, logistics bill 

framework and so on. Next, we will give a simple application 

case. First of all, the RPA robot will automatically log in to 

the logistics system, and use the logistics information capture 

framework to grab and identify the order information in the 

system. When a new order appears, it will automatically query 

the relevant information of the order in the system, as shown 

in figure 16. the query process will automatically enter the 

order information captured before the query input box, and 

automatically download the SKU file after querying the 

corresponding order. 

 

 
Figure .16 order information query 

 
Figure .17 SKU file download 

After downloading the SKU file, the RPA robot will 

automatically log in to the ASN system, upload the SKU 

according to the order and generate the freight list. As shown 

in figure18, the logistics document framework will be used to 

identify the SKU file and other documents, and the identified 

information will be sorted out automatically. At the end of all 

processes, the mail automation framework is invoked to 

notify the user by mail. 

 

 
Figure .18 Download of SKU file 

4.2.2. Automatic entry of manifest 

When the freight forwarder or shipping company or airline 

receives the customer's booking power or related documents, 

the customer service operator needs to enter the booking 

information into his own booking system. PA robot can use 

computer artificial intelligence technology to automatically 

scan and process the data of booking documents in various 

formats, and then input them automatically. When an 

abnormal situation occurs, the relevant personnel will be 

notified by email for manual inspection and review, and the 

manual input workload of this process will be greatly reduced. 

The specific process is shown in figure 19, this process 

mainly uses the logistics document framework to identify the 

power of attorney and read the information, the mail 

automation framework automatically sends mail to the 

business person, and the logistics information capture 



 

73 

framework reads the data of FMS. 

 

 
Figure .19 Automatic process for manifest entry 

4.3. Automation of delivery and sorting 

In the stage of shipping process, the RPA robot can 

regularly query the upcoming goods in the ERP system 

according to the number of the logistics order and 

automatically generate the shipping order, obtain the 

information of the goods to be shipped in the ERP system and 

the list of goods to be shipped in the goods delivery plan in a 

fixed time period, and generate the logistics order, including 

the detailed transportation mode, recipient and address and so 

on. Then the RPA robot will automatically log in to the 

logistics supplier system and automatically generate logistics 

orders. Logistics operators reduce repetitive work, convert to 

paying attention to abnormal situations and adjusting waybill, 

at the same time, improve the accuracy and efficiency of the 

whole process, let logistics information be updated, and 

enhance customer experience. 

In the delivery planning and tracking process, the RPA 

robot can automatically execute the pick-up request process, 

check the process, report the internal logistics system and the 

status of shipment between portals, automatic shipping and 

other processes. According to the mail received, the RPA 

robot can extract the shipping details, record them in the 

scheduling system, and automatically upload information 

such as car time in the web portals of customers and operators. 

The logistics information capture framework and the 

logistics information input framework are used in the delivery 

and sorting process. The RPA robot regularly acquires the 

shipping list information in the logistics system according to 

the logistics order number, and will use the logistics 

information capture framework to change the attributes of the 

logistics list location in the system for information capture. 

To input the captured information into the waybill, this step 

needs to use the logistics information entry framework to 

input the receiving address, recipient, transportation mode 

and other information into the corresponding location of the 

waybill, in which each location requires the developer to 

redevelop. configure the specific location and its attributes, 

and finally automatically transmit the waybill of the input 

information to the logistics supplier's system. The operation 

process is shown in figure 20, and the development process is 

shown in figure 21, which mainly includes three parts: data 

capture, data processing and data input. 

 

 
Figure .20 Automatic process for manifest entry 

 

 
Figure .21 Development process 

4.4. Automation of logistics information 

management 

The RPA robot can replace the related work of the logistics 

operator. The RPA robot automatically logs into the logistics 

information query system every day, monitors the status 

according to the logistics status information, downloads the 

data at the end of the signed status, needs to follow up and 

continue to track the unsigned status, automatically checks 

and tracks the abnormal logistics information, and 

automatically downloads the status categories. After the data 

is downloaded, the RPA robot automatically inputs the 

information completion system. 

The RPA robot can replace the related work of the logistics 

automatic return personnel. The RPA robot automatically logs 

into the logistics return system every day, uploads the finished 

data through the excel form, automatically checks the 

abnormal form information, collates the normal information, 

and automatically checks the abnormal form instead of 

manual review. 

The RPA robot can replace the related work of the logistics 

customs declaration personnel. The RPA robot automatically 

logs into the to-do and done task list of the customs 

declaration management platform system every day, submits 

and views the logistics customs declaration information, 

automatically matches and verifies the information of 

multiple approval stages, automatically downloads the 

information that meets the customs declaration requirements 

through the background system, and automatically inputs the 

customs declaration audit receipt processing time information 

into the EXCEL form. Calculate the time spent in the whole 

process, and then automatically send an email to the relevant 

personnel for review, or send it directly to the RPA robot for 

review. 

RPA robot can help supply chain managers automatically 

log in to the website or system provided by logistics suppliers 

after the goods are delivered, automatically query and update 

the logistics status according to the number of logistics orders, 

and input the latest information into the designated system, so 

that customers can know the current logistics status in time. 

The general steps of the above process are shown in the 

figure. First, the logistics web portal framework is used to 

automatically log in to the supplier's TMS system, then the 

logistics information capture framework is directly used to 

capture the information shown in the figure, and the mail 

automation framework is used to automatically send the 

information needed by the customer to the customer mailbox. 

Finally, the relevant information is automatically entered into 

the customer system through the logistics information entry 

framework. Complete the main process of logistics 

information management. 

 

 
Figure .22 Development process 

 

 
Figure .23 Development process 



 

74 

4.5. Automation of transport management 

The management process of transportation information is 

the same as the logistics information management process in 

the previous section, which is directly for secondary 

development. For the transportation of dangerous goods, the 

RPA robot can automatically declare before the transportation 

of the goods, automatically log in to the road transport safety 

supervision system, fill in the scheduling, consignment and 

other information in batches according to the information in 

the waybill application form, and generate declaration forms. 

The RPA robot completes the declaration process accurately 

and quickly, which can save a lot of time cost and labor cost. 

The dangerous goods transportation process mainly uses 

the logistics information input framework, in addition to the 

command to input text in the target, it also needs to use the 

commands such as setting element selection and setting 

element check shown in figure 24. In the secondary 

development, you only need to configure the selected element 

attributes to complete the transportation of dangerous goods 

waybill, as shown in figure 25. After filling in, the RPA robot 

will complete the relevant declaration process in the system 

and automate the whole transportation management process. 

 

 
Figure .24 Secondary development configuration 

 

 
Figure .25 Automatically fill in the relevant information 

4.6. Automation of financial issues in logistics 

management 

RPA robot can integrate data from logistics system, waybill 

payment system and other systems, and automate the process 

from order to cash for major 3PL operators. Solve the 

problems such as untimely time, inaccurate data, multi-

system information congestion and so on. PA robot can 

automatically obtain financial data in transportation, and at 

the same time use artificial intelligence, OCR and other 

technologies to scan invoices, instead of a large number of 

manual inputs, to achieve invoice automation in the logistics 

process. 

This process mainly uses the Mage AI module and OCR 

module in the logistics document framework, such as 

configuring the relevant commands shown in figure 26 in the 

framework to complete the relevant operations related to the 

invoice, in which the last line of command can carry out 

specific selection and configuration; you can choose different 

invoice types, you can choose to extract different information. 

 

 
Figure .26 Relevant commands for invoice identification 

4.7. Automation of logistics bidding 

RPA will log in to the bidding system or website provided 

by Party An on a regular basis, further check the bidding 

information of the project according to the set business logic, 

and then automatically upload the relevant bidding 

documents and automatically calculate the quotation. Solve 

the heavy workload, high repeatability, prone to errors, the 

time required to complete is very short, it is difficult to submit 

relevant information in time, the specific process is shown in 

figure 27. 

In this process, the logistics information capture framework 

is mainly used to capture the bidding information of logistics 

projects in the customer bidding system, and the logistics 

information input framework is used for automatic bidding. 

finally, the mail automation framework is used to 

automatically send mail to customers and responsible persons' 

mailboxes. 

 

 
Figure .27 Automation process of logistics bidding 

5. Conclusion 

Based on the business pain points in the logistics 

management process, this paper designs a RPA robot 

framework for logistics management. Developers carry out 

the second development on the basis of this framework, which 

can simplify the development process, improve the efficiency 

of development, save project preparation time, and improve 

the efficiency of logistics operation and management. The 

core module of the framework includes execution process 

module, initialization module, data processing module and 

exception handling module. The execution process block 

contains various functions of logistics management process, 

including logistics web portal framework, logistics 

information capture, input, logistics documents, mail 

automation and other frameworks. Finally, the designed 

Robotic Process Automation framework is applied to the 

logistics management process to verify the effect of the 

optimization. It is used in inventory management, order 

management, delivery and sorting, logistics information 

management, transportation management, financial problems 

in logistics management, logistics bidding and other 

processes. 



 

75 

References 

[1] Chen Yujie, Liu Xuejun. Research on the digital transformation 
mechanism of logistics enterprises' business processes under 
the normalization of epidemic prevention [J]. Business 
economic research, 2022 (02): 118-121. 

[2] Cheng Ping, Dai Jia. Research on general ledger management 
optimization of Financial Shared Service Center Based on RPA 
-- Taking ABC Group as an example [J]. Friends of accounting, 
2018 (16): 141-146. 

[3] Cheng Ping, Tao Siying, Zhu Bofeng. Research on financial 
report generation optimization of Financial Sharing Service 
Center Based on RPA [J]. Friends of accounting, 2018 (22): 
146-151. 

[4] Cheng Ping, Liu Hongju. RPA based financial sharing service 
center account table reconciliation process optimization [J]. 
Friends of accounting, 2018 (21): 145-149. 

[5] Cheng Ping, Wang Jianjun. Research on the optimization of 
Accounts Payable process of Financial Sharing Service Center 
Based on RPA [J]. Friends of accounting, 2018 (19): 154-160. 

[6] Cheng Ping, Wang Shuang. Research on fixed asset 
management optimization of Financial Sharing Service Center 
Based on RPA [J]. Friends of accounting, 2018 (20): 148-153. 

[7] Cheng Ping, Ji Wei. Research on Optimization of accounts 
receivable management of Financial Sharing Center Based on 
RPA [J]. Friends of accounting, 2018 (15): 153-157. 

[8] Cheng Ping, Zhang Hongshuang. Research on tax management 
optimization of Financial Sharing Center Based on RPA [J]. 
Friends of accounting, 2018 (14): 145-148. 

[9] Tian Gaoliang, Chen Hu, Guo Yi, Xue Yuting. Research on the 
application of financial robots based on RPA technology [J]. 
Journal of Finance and accounting, 2019 (18): 10-14. Doi: 
10.19641/j.cnki.42-1290/f.2019.18.002. 

[10] Quan Wenju. Practice of power enterprises to assist financial 
digital transformation based on RPA technology [J]. 
Accounting study, 2019 (30): 15-16. 

[11] He Hong, Liu Xiyu. Analysis on the application of RPA in the 
financial field of insurance enterprises [J]. Journal of insurance 
vocational college, 2018,32 (05): 62-65. 

[12] Liu Qingying. Influence and promotion of financial robot on 
enterprise financial management [J]. Business information (44): 
6. 

[13] Lu Ying. Exploration of intelligent financial application 
scenarios of real estate group [J]. Chinese chief accountant, 
2019 (04): 98-99. 

[14] Jiang Mingyu Design and application of uipath based robot 
process automation framework [D]. Shandong University of 
science and technology, 2020.DOI:10.27275 / 
d.cnki.gsdku.2020.000853. 

[15] Yu Yanling. Discussion on enterprise logistics marketing 
combination mode under the digital background [J]. Logistics 
technology, 2012,31 (15): 53-55. 

[16] Yu Lei. Internet of things logistics management system based 
on RFID electronic tag [J]. Microcomputer information, 2006 
(02): 233-235 + 232. 

[17] Gao Yanfeng. RPA, financial intelligence automation reshapes 
the financial function of enterprises [J]. Accounting learning, 
2019 (35): 55-56..