Digital transformation has a long way to go, so how to take the key first step

The manufacturing industry has a huge system 、 Long business chain, etc , There are many aspects involved in the digital transformation . Many enterprises have no way to start in the face of digital transformation , How to take the key first step , Realize the rapid realization of data assets ？

The development trend and opportunities of China's manufacturing industry

Manufacturing is the foundation of the real economy , It is the key to high-quality economic development in the future . In the global competition facing the digital economy era , Rely on digital technology to develop to a higher level 、 More competitive advanced manufacturing , It has become the strategic consensus of all countries . In recent years , Germany put forward “ Industry  4.0” planning , The United States put forward “ National manufacturing innovation network ”, Japan put forward “ Innovative industrial structure plan ”, China has also proposed “ Made in China 2025” development planning , The common ground is to make full use of the Internet of things 、5G  signal communication 、 robot 、 Artificial intelligence and other technical means improve the intelligence of the manufacturing industry 、 Unmanned degree .

In this context , As a manufacturing power , China's manufacturing industry has also ushered in new development . According to the data , By the end of 2021 year , The added value of China's manufacturing industry accounts for GDP Share to 27.4% Year on year increase 1.1 percentage ,2021 It also ranked the first in China's manufacturing added value in a row 12 Ranked first in the world .

While China's manufacturing industry is booming , Some manufacturing outflows have also occurred . Weakened by some cost advantages 、 Driven by factors such as policy decline , Some medium and low-end industries flow to Vietnam 、 Indonesia and other Southeast Asian countries , Superimposed on the current escalation of domestic epidemic control , It has a certain impact on the supply chain of some industries .

meanwhile , The development of China's manufacturing industry has also encountered some bottlenecks . At present, some industries are still at the middle and low end of the global value chain , The added value of products is low . For example, there are chips in China “ thetime ” The phenomenon of , In addition, in the aviation industry 、 The integrated circuit 、 High end CNC machine tools 、 Agriculture mechanical 、 There is still a certain gap between the core technologies in the fields of high-performance medical machinery and developed countries , Lack of core competitiveness . At the same time, enterprises pay less attention to environmental protection , There is “ high energy-consumption 、 High emission 、 High pollution ” The phenomenon .

How to take the first step in the digital transformation of manufacturing enterprises

Digital transformation of manufacturing industry , It is the business transformation under the guidance of strategy , It's data driven 、 Intelligent assisted research and development 、 production 、 operating 、 Service improvement , Finally, we will promote the optimization of profit model and the improvement of user experience . It needs to promote the two-way integration of business and system , Take digitization as the core , With the help of network means , Realize intelligent empowerment , Ensure that products and services are delivered with high efficiency and quality , Continue to enhance the core competitiveness of enterprises . The three core points are to achieve strategic change , Improve the efficiency of enterprise operation , Enhance user experience .

How to use the data , Data driven business , The first step in the digital transformation . However, the manufacturing industry has common pain points in data application ：

1、 data silos

Manufacturing enterprises have a long link from production to sales 、 There are many links , Data exists in different systems 、 The data caliber is inconsistent 、 Data granularity is coarse 、 Low update frequency . Internal and external data of the enterprise are scattered , The caliber and quality of the data are inconsistent , As a result, when business personnel analyze data , It takes a lot of time to integrate and clean the data in different systems , Even a large number of missing data , The final result lacks accuracy .

2、 Insufficient data application

The digitization of manufacturing enterprises started relatively late , There are few applications of data in actual business , Business value has not been fully verified . The application of data in enterprises is mostly simple data collection 、 Statistics 、 contrast , There are still a lot of deficiencies in deep excavation and insight into the data , Failed to compare the data analysis method with the actual business scenario decision 、 Process integration , Lack of data application practice cases .

3、 Data value inhibition

Multiple factors cause the data value of the enterprise to be suppressed , Difficult to guide the business with data .

To address these issues ,BI Because of its simplicity and ease of use 、 Quick liquidation and other features have become the focus of major enterprises .BI Its advantage is that it can integrate data from different sources to solve data islands , You can use BI The data analysis method and data processing technology in the system enable business personnel to effectively combine business experience and data to form data applications , In addition, various reports and Kanban can be provided accurately and quickly through analysis and query tools , Provide decision support for enterprises , Bring the value of data .

utilize BI Realizing data-driven business growth is usually divided into  4  Stages ： Result monitoring 、 Problem diagnosis 、 Decision support 、 Intelligent prediction .

1、 Result monitoring

Timely and accurate monitoring of current business objects , It can give early warning and prompt in case of abnormal conditions . At this level , We need to constantly look at the outcome indicators ; Help monitor business status in a timely manner , Improve regulatory efficiency , Realize intelligent operation .

Take the visual analysis of a mobile phone manufacturing industry as an example . adopt BI Monitor the production of the workshop , Build indicators , Know the plan achievement rate 、 Whether the production line is normal 、 Whether the output of each workshop is up to standard , Through such monitoring, the leaders of the factory can directly know the current situation of production , What are the problems .

2、 Problem diagnosis

Just looking at the results is not enough , You also need to quickly find the business problem . Combined with scenario business logic , And data , Through multidimensional analysis , Layer by layer insight into data , Quickly locate business issues . There are also preset rules , Highlight the overdue business phase . Help quickly trace the source of the problem , Improve the efficiency of problem diagnosis .

Take a home appliance enterprise as an example . Product delivery time is an important reflection of an enterprise's production capacity , It is also an important indicator that production managers pay most attention to , At the same time, because of its complex influencing factors , And let the manager have no way to start . The delivery date of the enterprise is determined by the average delivery date 、 Delivery time compliance rate 、 The average number of affected days and the number of non-conforming orders constitute . When the average delivery time is not ideal , It can be used to diagnose the problem , It is found that the average delivery time becomes longer due to the instrument production line , The instrument production line is caused by the long process of financial audit , So as to trace back to the root causes that affect the product delivery time , Help enterprises take more targeted measures , To improve product delivery efficiency .

3、 Decision support

Finding the problem is not the fundamental goal , The fundamental goal is to find a solution to the problem , Help decision support . What do you mean “ Real decision support ”？ That is, the decision-making method can be found directly from the data .

Take a home furnishing enterprise as an example , From the perspective of the purchasing manager , What I care about most at work every day is ：

l   Which models / Materials need to place purchase orders today ？

l   To which supplier ？

l   How much is it ？

l   How likely is the supplier to delay delivery ？

Subject to the goods “ Take simple cloth sofa in northern Europe as an example ”, When you click this kind of product, linkage analysis will be performed . Discover through inventory change trend , If not replenished , The first 7 The inventory of simple cloth sofa in northern Europe will fall below the safety stock value . In a similar way , That would answer “ Which models / Materials need to place purchase orders today ？“ This problem .

The next thing to look at is which supplier ？ How much quantity and how likely is the supplier to delay delivery .

Assume that the order quantity is 500, Through the supplier information, we can find , From the perspective of minimum order requirements ,A/B/C All three suppliers are satisfied with ; Alert by highlighting the red light , Further discovery of suppliers A The current in the “ busy “ The state of , The need to 7 Supply within days , Therefore, from the perspective of supplier delivery safety , Can exclude A.

From the perspective of historical delivery accuracy of suppliers , Preference B. If the premise is to have greater inventory redundancy , such as 700 The number of , Then some of the quantity is not so urgent , From the perspective of supplier balance , Think about 500 to B, Put the rest 200 Next to the C.

From the perspective of the purchasing manager , It turns out that you can check all kinds of form data and even call to confirm one by one , As soon as possible 20、30 minute , Now through BI use 1 Minutes to complete , This means that the decision-making efficiency has been improved dozens of times than before .

4、 Intelligent prediction

In addition to using current or historical data to help find problems in the current business , Find the cause , Assist leaders to make decision support , Historical data can also be used to establish AI The model predicts the future , For example, in the manufacturing industry, predict which equipment may have faults .

Take a fan enterprise as an example . Generally speaking , The service life of the fan is 20 year . Most fans are guaranteed for two to five years of initial operation , And the maintenance mode of the fan is mostly “ Passive operation and maintenance ” Mainly , It mainly relies on site staff for regular maintenance and troubleshooting . Simply rely on manual maintenance , The cost of operation and maintenance is very high , It is also prone to power generation losses and even operation safety problems caused by different levels of personnel .

In this case, we hope to establish a prediction model , Help find the fault of the fan in time . First, determine the requirements , Then collect the data , Do data cleaning 、 Preprocessing 、 Feature Engineering , The use of AI Algorithmic modeling , Deploy the model , Finally, the accuracy of the model can reach 80%  about . With such intelligent prediction , You can know which fans have faults in advance , Targeted maintenance , Reduce the cost of people and time .

Typical cases ： The first step in the transformation of global household appliance manufacturing enterprises

This enterprise is a benchmarking representative of digital transformation in manufacturing enterprises . As early as 2012 year , The enterprise began its digital transformation , So far, it has entered the 6 Stages . Each stage of transformation is to solve a specific problem , The challenges encountered at each stage are also different .

The company started in 2015 In, it reached cooperation with Yonghong technology , Yonghong BI As the enterprise's digital transformation 1.0 An important part of the project . The essence of digital construction in this period is the establishment of information system , meanwhile BI The platform can be used as a supplement to the information system , The two are interdependent , It can be realized IT Business data , It can also be realized IT Data informatization , The two drive each other .

The strategy of the enterprise is to make digital management transparent first , Post data driven optimization management . Unified data caliber for business , Build a business indicator system , By mobile phone 、 The computer 、 The large screen Kanban can understand the operation of the business unit in real time , Find problems through big data , Drive business optimization 、 Management closed loop . In terms of technology and personnel, the data of the whole value chain is connected , Build a big data platform , Improve the big data team .

The business pain point of the enterprise is that it cannot see the market situation clearly , Do not understand the market capacity 、 Competing brands 、 Main selling price , Business decisions exist “ Beat the head ” The situation of . Store operation depends on experience , The management of the store is not enough , Unscientific operation evaluation . The product performance is not clear , Do not know the best-selling products 、 New product performance 、 Reasons for the loss of users , It is difficult to accurately increase sales . The voice of the user cannot be heard , Do not understand brand reputation 、 Management of user complaints and negative comments , Difficult to improve brand reputation 、 Targeted improvement products .

IT The pain is that , The original tradition BI The platform encountered difficulties in report development 、 Problems such as the inability to implement authority control and the delay in responding to business needs .2015 Agile in BI When selecting the model , Some of the core concerns include ：

l   Be able to respond quickly to requirements , As a supplement , Then gradually replace the traditional BI;

l   High performance , It can meet the group's big data computing requirements ;

l   Ease of use and customizability , On the one hand, let the business departments conduct data analysis by themselves , On the other hand, it can meet the group's requirements for customization ;

l   It should have a perfect authority management mechanism , It can adapt to the characteristics of multi organizational groups .

For business problems , The enterprise has built a market-oriented 、 Three modules of data application for users and internal operation of the enterprise .

1、   Market data analysis

Collect five e-commerce data , Insight into the market pattern , Analyze the market structure of the industry from the perspective of the Internet 、 Development trend and user reputation , Help enterprises to deeply analyze the market 、 Analyze users 、 Product and competitor analysis , Continuously optimize the existing product layout , Promote product improvement .

2、 User data analysis

Create a portrait of thousands of users , Integrate internal and external data , Build a user panoramic view , formation 2 Billion unique identity users ,10 Billion + Behavioral data ,800+ Label type , Provide accurate support for various promotional activities .

3、 Internal operation data analysis

Its goal is to hope that the enterprise can understand its own situation more thoroughly , Build a business analysis center for the group , Create unified data , A unified platform , Unified operation to achieve agile business insight , Including finance 、 marketing 、 Supply chain 、 Human resources, etc. , Accumulate data , Improve efficiency , Release value .