5 Challenges Holding Back Smart Factory Further Development

With the popularization of big data, the upgrade of networking conditions, and the rapid development of computer processing power, artificial intelligence and Internet of Things technology will become the key to accelerating the transformation of factories. But according to a report released by Intel Corp., two-thirds of the manufacturing companies that have applied digital technology have not yet implemented such projects significantly. What is the reason?

Intel has conducted a two-year survey on more than 400 manufacturing companies and technical experts for the development of intelligent technologies, solutions or services to explore the key elements for the implementation and popularization of Industry 4.0. The survey, which just released the second part of its report, "Accelerating Industrialization," focuses on how smart factory workers can adapt and respond to the role of artificial intelligence in manufacturing, and how leaders can take this further through corporate strategy Accelerate this transformation process.

While 83% of companies plan to invest in smart factory technology within the next two to three years, they are often unsure how to proceed, or are hesitant to face the associated risks, the study found. So what are the challenges that make smart factory projects difficult to implement or scale up? What can leaders do to avoid the pitfalls? Below are the top five challenges raised by respondents, along with recommendations from Intel.

Challenge 1: Technical Skills Gap

36% of respondents said that technical skills gaps are a hindrance to companies' investment in smart factory projects. To successfully implement new technologies and maintain operations, a company's workforce must be "digital agile": employees must understand manufacturing processes and master the digital tools that support those processes.

Solutions:

· Create courses to support ongoing learning for existing staff, then combine new concepts with first-hand practical opportunities to enable staff to work on operational activities applied in the context of ; building interconnected modules that allow employees to gradually develop and hone their skills to increase efficiency.

· Provide guidance on digital tools and skills that may increase in importance in the future. The content of the materials should be comprehensive, including network security, infrastructure, artificial intelligence, data, storage and computing power requirements. It explains various concepts and how they relate to each other.

· There should be more emphasis on problem assessment and resolution than on solution implementation.

· When launching new smart technology projects, strike a balance between hiring external experts and developing in-house talent so that the company's digital capabilities are sufficiently developed.

Challenge 2: Data Sensitivity

27% of respondents mentioned "data sensitivity" and expressed more concern about issues such as privacy, ownership and management of data and IP. more and more attention.

For example, the successful deployment of an AI algorithm requires large amounts of data to be trained and then tested. This means data must be shared, but many companies are reluctant to share their data with third-party solution developers. Furthermore, respondents generally felt that existing data security policies within the organization were insufficient to support such sharing activities.

Solutions:

· Normalize the data sharing policy, so that the data transfer inside and outside the organization can be carried out smoothly

· Develop data governance policies to balance trade-offs Sharing data and the risks that come with it. However, it should also be noted that the formulation of policies must be adjusted according to specific circumstances, and the practice of generalizing them will leave hidden dangers. If possible, relevant clauses should also be included in future contracts with suppliers.

· Consider data sharing needs in advance of a business intelligence project and allow time in the project operations schedule to negotiate such needs.

Challenge 3: Interoperability

23% of respondents said that the protocols, components, products and systems used in the enterprise lack interoperability. This kind of trouble has always existed, but it is especially critical in the process of digital transformation. The lack of this aspect limits the ability of enterprises to innovate and upgrade system components, because they cannot easily change suppliers or replace the system components. a component of .

Solutions:

· Actively promote and support the development of relevant standards to improve interoperability; participate in relevant associations and groups as much as possible.

· Collaborate with vendors to develop and deploy modular solutions and seek to advance multi-vendor approaches to open up long-term upgrade paths.

· Consider open source solutions when launching smart technology projects.

Challenge 4: Safety

22% of respondents mentioned security threats and expressed concerns about existing and potential future hazards in the factory.

Smart factories combine physical and digital systems, which greatly facilitates the real-time operation of interconnected systems by employees, but also increases the risk of hacking. The network used in the smart factory is connected to a large number of machines and devices, and the vulnerability of any one machine can be exploited to attack the entire network. Organizations need to predict vulnerabilities at the operational level of enterprise systems and specific machines. However, the defense means of enterprises are generally insufficient, and they often rely on technology and solution providers to do it for them.

Solution:

· Involve technical talent in operations and IT on intelligent project teams to assess possible vulnerabilities and weaknesses in people, processes, machines and networks

p>

· Learn about new trends in suppliers' equipment and operations, and predict what new vulnerabilities may emerge

· Rehearse for some extreme cases, such as every None of the components have major vulnerabilities, but linkages with each other create new vulnerabilities. Be prepared for this situation.

Challenge 5: Dealing with the Growth of Data

18% of the respondents mentioned the growth in the size and speed of data generation in their enterprises, as well as the mechanisms for utilizing the data. With the application of artificial intelligence, enterprises will generate more data at a faster rate, and the data format will be more diverse. Therefore, the difficulty of applying data has also increased exponentially. For example, recommendations made by artificial intelligence must be able to combine data generated in different ways and at different frequencies. On the other hand, AI algorithms must be easy to understand in order to be helpful in decision-making.

Solutions:

· To understand the data that can generate business value, balance computing power, bandwidth and real-time (low latency) feedback from the level of enterprise assets demand.

· To be able to predict in advance a sampling rate sufficient to reflect the state of the machine or operation. It is not necessary to collect all data.

· Establish a robust system architecture prior to the implementation of intelligent projects to balance computing power needs with where they are generated (for example, consider edge or cloud computing), current and future storage needs to build communication infrastructure.