Complex businesses have complex operations. In this video, Huawei’s Head of Intelligent Production Scheduling and Optimization explains how managing hundreds of factories manufacturing thousands of products requires handling millions of variables.
By using mathematical models and algorithms, Huawei works toward finding optimal solutions that balance multiple goals in a quest for optimization inside the company. This quest and the ultimate solutions create value for Huawei – and the lessons learned in-house can be applied to the operations of customers as well.
Huawei innovations in intelligent operations are being applied across diverse scenarios such as airports, power grids, ships, ports and factories to provide a boost to productivity.
Watch the video below to discover how Huawei is automating decision making to optimally allocate limited resources,helping customers become more efficient, cut costs or both.
Q:
Tell me a little bit about yourself.
Dan Zhenggang:
I'm Dan Zhenggang. I got my bachelor's degree and doctorate degree from Tsinghua University. Since joining Huawei, I have worked on the research and algorithm design and planning for supply chains. The physical world around us has limited resources. Considering this, how can we use these resources to achieve the best possible results? This is the focus of my research. We need to consider all available resources and all real world scenarios to create a mathematical model. Then you must find the right formulas to solve that mathematical model. In this way, we can find the mathematically optimal solution.
Q:
What's the biggest impact on people's lives that you think your research will have?
Dan Zhenggang:
Operations optimization creates value in many respects. A company doesn't need to expand its facilities or buy more equipment, but the company itself or one of its plants can increase production capacity by 10% or even 20%. You can work much faster when computers are helping you make decisions. Moreover, our world will become cleaner and greener. For example, manmade resources, like trains and aircraft, are limited. How can aircraft transport more passengers, while using less energy? This is about doing more with less. At an airport, a great many resources must be managed. To do this efficiently, we must create the optimal resource scheduling plan. We need to make sure that as many aircraft as possible can park at jet bridges, so that passengers can directly board planes.
We need to reduce the time each aircraft spends at a jet bridge. We must make sure that passengers board as quickly as possible. These matters require careful consideration and optimization. There is another key challenge faced by airports. Abnormalities are not uncommon. For example, blizzards, storms, and heavy rains. When any of these occur, things can get messy, and we need to adjust a contingency plan. If this plan is adjusted manually, something might be missed, because many things need to be factored in. Besides, manual decision making may take a long time. Our hope is that, when an emergency occurs, we can get everything back on track rapidly through operations optimization.
At Huawei, there are tens of thousands of products in production. In the Pearl River Delta in China, we manage several hundred plants. This means several million variables are at play. In my daily work, I have to find solutions within hours. Such solutions specify the plants' production mode for the next day or for a certain period of time. These solutions show how processing orders can be released to the plants. During this process, these solutions create value for us. They help lower the costs of both raw materials and finished products. They also support the rapid shipment of products to customers, meaning products don't sit idle in warehouses.
Q:
Can you tell me about some of the biggest difficulties you've faced?
Dan Zhenggang:
One of the biggest difficulties is understanding customers' business processes, like those of airports and steel plants. We have to dive into these processes. Another challenge is that we must convert business problems into an abstract math model. The next step is solving the problems. This may involve huge amounts of data. We may need to solve problems using millions of formulas. It all comes down to finding optimal solutions that balance multiple goals. This is a complex undertaking. Our technologies are already being trialed in many industries, like ports, airports, transportation, logistics, manufacturing, automotive, wine warehousing, and steel making. We already have many mature applications and successful experience. We can share this experience with customers in these industries. We hope to help customers improve efficiency, cut costs, or both. We hope that our technologies will see wider adoption.
Q:
So as you know, Huawei are committed to a fully connected intelligent world. In the scope of your research, what do you see that world to be like?
Dan Zhenggang:
To build a fully connected, intelligent world, we must provide end-to-end technology to solve problems and support the entire world. In a sense, the world around us is disorderly, full of irregularities and randomness. Addressing this disorder requires the connection of all types of data and continuous pursuit of optimization. So our technology provides the means to achieving optimization. This is how my work creates value.