On the evening of November 17, the fourth phase of the “AI + Logistics Beacon Live Room” series of live broadcasts jointly created by “Logistics Technology and Application” magazine, the Artificial Intelligence Logistics Industry Alliance and Megvii Technology was launched online on time.
(ChinaIT.com News) As the guest of this live broadcast, Professor Zhao Ning, a doctoral supervisor of University of Science and Technology Beijing and the executive editor of “Logistics Technology and Application”, made an in-depth sharing on the topic of “Digital Twins and AI Logistics”, combining multiple digital factories Simulation planning projects, so as to better help manufacturing and logistics industries to apply logistics simulation technology.
Professor Zhao Ning is engaged in the teaching and research of manufacturing system and logistics system scheduling and simulation technology in the Department of Logistics Engineering, School of Mechanical Engineering, University of Science and Technology Beijing. He has participated in the digitalization of many large enterprises such as Siemens, Mousse Bedding, Jiumu Kitchen and Bathroom, Han’s Laser, etc. The factory simulation planning project is committed to promoting the widespread application of logistics simulation technology in the manufacturing and logistics industries. At the same time, he is also a member of the Intelligent Simulation Optimization and Scheduling Special Committee of the Chinese Society of Simulation, a member of the Industrial Big Data and Intelligent System Branch of the Chinese Mechanical Engineering Society, a member of the Working Group of the Logistics Management and Engineering Professional Teaching Steering Committee of the Ministry of Education, and China Logistics The sixth special researcher of the society.
Professor Zhao Ning’s speech is rich and comprehensive, mainly including three parts:
1. The emergence and development of concepts such as digital twins, AI logistics, and intelligent scheduling.
2. The principle basis and development status of the application of digital twin theory and technology in the field of intelligent logistics.
3. Introduce the application of digital twin and simulation technology in multi-layer shuttle system scheduling and KIVA robot system scheduling research, as well as the scientific research results obtained.
According to Professor Ning Zhao, the concept of Digital Twin was first proposed by Professor Grieves of the University of Michigan in the Product Lifecycle Management (PLM) course in 2003, but its theoretical concept was widely disseminated, thanks to the 2011 National Aeronautics and Astronautics Corporation. The Bureau explores the use of (Digital Twin (DT) technology for the health maintenance and protection of aerospace vehicles. First, establish a model of the real aircraft in the digital space, and achieve complete synchronization with the real state of the aircraft through sensors, so that after each flight, according to The existing conditions and past loads of the structure are analyzed and evaluated in time to determine whether maintenance is needed and whether it can withstand the load of the next mission.
A digital twin is a digital representation of the physical world. In addition to aircraft operation and maintenance, it can be used in many fields such as product design, manufacturing, urban management, logistics, etc., which has attracted widespread attention from the academic community.
Professor Ning Zhao continued to interpret the concepts of intelligent (AI) logistics, intelligent scheduling, and logistics simulation. Specifically: AI logistics is the use of integrated intelligent technology to enable the logistics system to imitate human intelligence, with the ability to think, perceive, learn, reason and judge and solve certain problems in logistics by itself. The future development of intelligent logistics will reflect four characteristics: intelligence, integration and hierarchy, flexibility, and socialization. Intelligent logistics is realized through the four intelligent mechanisms of the intelligent logistics system, namely, intelligent information acquisition technology, intelligent transmission technology, intelligent processing technology, and intelligent application technology.
Intelligent scheduling, also known as knowledge-based scheduling, is one of the research fields of interest in artificial intelligence and intelligent control. Many combination problems in reality are more complicated. To find an optimal scheduling scheme from possible combinations or sequences requires a large search space, which may result in combinatorial explosion problems. Intelligent scheduling is a scheduling method that makes full use of the knowledge of the problem domain to reduce the combinatorial explosion as much as possible, so that the optimal scheduling or combination problem can be effectively solved.
Logistics simulation is to model the logistics system, and compile corresponding application programs on the electronic computer to simulate the actual logistics system operation status, and to count and analyze the simulation results to guide the planning, design and operation management of the actual logistics system. In reality, many logistics systems are relatively complex, and simulation is needed to enhance understanding and optimize control. The simulation principle is discrete event simulation.
Later, Professor Zhao Ning also analyzed and clarified several similar technologies for everyone, including digital twins and animation, digital twins and simulation, digital twins and virtual scheduling, digital twins and virtual monitoring, and so on.
In general, these technologies are similar to digital twins, but there are differences. He explained that, as opposed to animation, digital twins are not only a mirror image of the external physical world, but more importantly, they mimic the performance of foreign logistics equipment or systems. The value of virtual debugging is that low-cost computer experiments can be used to replace high-cost physical experiments, but the difference between the two is that physical simulation is generally oriented to the system, while virtual debugging is generally oriented to equipment.
Virtual debugging is more pursuing “true”, the purpose is to test logic to avoid errors; if it is not “fast”, it is difficult to optimize at the system level. The ultimate “truth” is often the result of multi-discipline and multi-field integration. The ultimate “fast” requires a deeper understanding of mathematics and data.
Speaking of the difference between digital twin and virtual surveillance, Professor Zhao Ning explained that with the popularization of 3D animation technology, virtual surveillance technology that uses virtual simulation animation instead of camera to shoot live video has gradually appeared. Virtual monitoring is driven by on-site real-time data, displaying real-time data in the form of animation; while virtual simulation is writing the system’s internal logic into the simulation model, and relying on the system’s internal logic to drive simulation and animation.
In the speech, Professor Ning Zhao focused on the application and development of digital twin technology in intelligent (AI) logistics. He introduced that there are two main technical lines for the development of machine intelligence, one is artificial neural network and the other is intelligent algorithm.
Artificial Neural Networks (ANNs) are also referred to as Neural Networks (NNs) or Connection Models for short. It is an algorithmic mathematical model that imitates the behavioral characteristics of animal neural networks and performs distributed parallel information processing. This kind of network relies on the complexity of the system, and achieves the purpose of processing information by adjusting the interconnection between a large number of internal nodes.
“Intelligent algorithm” refers to some relatively “novel” algorithms or theories in engineering practice, such as simulated annealing algorithm, genetic algorithm, differential evolution algorithm, particle swarm algorithm, ant colony algorithm, etc. These algorithms or theories have some common characteristics (such as simulating natural processes), which are very useful in solving some complex engineering problems.
Compared with digital twin technology, Professor Zhao Ning believes that virtual scheduling and simulation technology have some difficulties when applied to intelligent logistics systems.
Among them, the main difficulty of the virtual scheduling application is: the scheduling model cannot fully consider all real factors, such as the congestion problem of multiple robots, the deadlock problem, and so on. Dynamic information is difficult to obtain, such as robot fault information. Optimization goals are difficult to determine, such as how to choose between the highest efficiency, the shortest path, and the highest hit rate. Constraints are difficult to determine, for example, empty robots can wear shelves but not columns; one-way/two-way streets, etc. It is difficult to evaluate scheduling algorithms. For example, are well-applied algorithms in 10 robotic warehouses still effective in 100 robotic warehouses?
The main difficulty in the application of simulation technology is: the application goal is not clear, except for animation, what problems can be solved? The cost is high, the simulation modeling is fully customized, the cycle is long, the labor is consuming, and the cost is high. The degree of fidelity is low, and the deviation between simulation results and physical results is large. The simulation speed is slow, and the simulation speed is close to or even slower than the physical time.
If there is an ultimate way to solve the above problems, it is a digital twin for AI logistics.
In the scientific research work, Professor Zhao Ning has carried out research on the optimization of multi-layer shuttle system scheduling and KIVA robot system scheduling using digital twin and simulation technologies, and has obtained relevant scientific research results.
Finally, Professor Ning Zhao made an outlook and prediction on the future development direction and related issues of digital twin technology and smart logistics:
- Intelligent logistics has brought great changes to the traditional logistics model, eliminating many pain points caused by the application of manpower, and at the same time bringing about new demands for the use of intelligent equipment.
- The integration of design, scheduling, and simulation of intelligent logistics systems is a promising application direction.
- The combination of smart logistics and digital twins is the direction of the future.
- The “stable, accurate and fast” scheduling and the simulation “true and fast” are the core of development, and have strong application value to the logistics equipment industry.
- Enterprises focus on landing, universities focus on frontiers, and innovation in the field of logistics equipment requires the deep integration of companies and universities at the level of research and development.
Articles on the ChinaIT.com website are limited to providing more information and do not represent the standpoint of this website. For reprint, please indicate the source. The reprinted article comes from the Internet. If you have any copyright issues, please contact us: content@chinait.com.
.