Amazon’s Kiva a Poor Match for China’s Logistics

Fang Yu / Journalist / 2015-01-27

Automation can be a good tool rather than a goal. Nowadays, building a solid information infrastructure and having warehouse process standardization is more important for the China’s e-commerce and logistics industry.

Read this article in Chinese

If there is someone in the United States busier than Santa Clause during Christmas season, it’s the online retailing giant Amazon. And its magic reindeer is the star robot of its warehouse—Kiva.

This short bright orange chunky machine is very intelligent, as well as strong. It’s able to load and move up to 750 pounds of cargo. It runs swiftly around the warehouse, fetching and delivering ordered goods, following computer instructions. But the most effective route is calculated by itself. Not only that, it can also automatically avoid obstacles and its “colleague” robots, turning and proceeding smoothly.

Photo credits: Kivasystems inc.

At the recent China E-Commerce Cross-Industry Summit in Beijing, a senior executive from Amazon China displayed a video of Kivas working between shelves in the company’s warehouse. In the video, a dozen busy Kivas worked very smoothly with only 1 or 2 human workers occasionally appearing on the screen. The video successfully captivated the audience. (A video from amazonpickingchallenge.com)

What is interesting though is that right after the video, another speaker poured cold water on the great passion aroused by the little robot. Mr. Zhou Tao, vice president for logistics management at the Chinese e-commerce company Jumei Youpin, said, “A Kiva costs almost 300,000 RMB. No one could afford that here!”

Zhou Tao’s words are worth noting, for he used to be the vice president of Amazon China and was in charge of constructing the Tianjin warehouse, the company’s largest warehouse in China. On one hand, most Chinese logistics practitioners admit that the low level of automation is restraining the development of the industry; logistics costs today still account for 15-20% of the total manufacturing costs in China, while the figure in the United States is in the single digits. On the other hand, robots like Kiva have been accused of being “too advanced” for local players.

While the debate about the prohibitive price of Kiva is misleading, in my opinion, one thing calls for deeper questioning: Is full automation enough to offer the way out for China’s stumbling modern logistics industry? To what extent can intelligent equipment and processes contribute to the optimization and efficiency of the whole system?

Make no mistake. I am not saying that technology is irrelevant in this case. Without any doubt, technology is the key factor in all industries and economies, especially from a long-term perspective. Tools and process innovations bring new efficiency and models, stimulating and accelerating industrial upgrades.

However, in contrast to the overwhelming praise Kiva has harvested so far, rarely are there any Chinese companies showing interest in introducing it to their own warehouses. Even Amazon China didn’t put Kiva to use on a large scale like it did in the US. Currently, it has 15,000 Kivas running around in Amazon’s American warehouses.

Chinese e-commerce enterprises are expanding their business to smaller cities and remote rural areas where the warehouses are dispersed and the infrastructure worse. As a result, automation has little to do. /Phonto: connecting4.com

The NASDAQ listed company Jingdong, Amazon’s counterpart in China, has established its first “Asia One” modern logistics center in Shanghai, which is around 200,000 square meters (equivalent to 14 football fields). Wang Yinxue, managing director of the project, believes that robots like Kiva and delivery drones will definitely change the industry in the future. However, whether full automation should be a priority for the company remains to be seen at the current stage.

Scale may be one concern. It is estimated that Amazon China’s total warehouse space is around 900,000 square meters, while the up-to-date figure for Jingdong is 1.5 million square meters. As more and more projects like “Asia One” in other big cities are expected to be established soon by Jingdong, the size of the two will become incomparable. Though Amazon’s warehouses and logistics management are generally more automated and intelligent, it is achieved within a limited scale. Furthermore, when it comes to the application in scale, the 300,000 yuan unit price of Kiva becomes an inevitable issue.

More importantly, automation is only part of the systematic work, which should be based on the accurate matching of information and cargo flows along the entire process of sorting, picking and delivering. Most of the Chinese e-commerce companies are still at the stage of building that fundamental information infrastructure. Automation in some specific functions (robotic picking arms for instance) is thus very limited in improving system efficiency.

Wang Yinxue pointed out that most of the products and solutions provided by foreign companies today are not capable enough to meet the Chinese logistics market needs, which are characterized by huge capacity and special customer needs. Take picking machines as an example. It is very difficult to find a product with the ability to handle 800,000 orders per day, or 40,000 per hour, which is an ordinary business volume for a large Chinese e-commerce company.

During online shopping peak periods like “Double Eleven,” overwhelmed by a flood of orders, the only priority for companies like Jingdong is to “send the things out.” In such a situation, the flexible brain and skilled hands of human workers are the most efficient solutions. Huge volume, limitless variety of goods, urgency of the demands—each of these can render a warehouse controlled by an automated system paralyzed in a few seconds.

To some extent, a high degree of automation only leads to slightly higher average working efficiency. For systems and processes with large volatility, being flexible and responsive to emergencies is the most important. Full automation here cannot be seen as an advantage anymore; rather, it’s risky. The more it is automated the less capable it is to respond to and deal with unexpected variables and interference from outside the system.

The logistics chain can be seen as a highway. Each perfectly designed highway is a huge closed system with beautifully organized entrances and exits, sideways and crossroads. The structure seems perfect, but in China when it is set free for use during holidays, without any exception, the highway will get paralyzed as millions of cars rush in.

That’s exactly what most delivery companies are faced with during “Double Eleven” in China. They have to make contingency plans and set up “guerrillas”—thousands of casual workers—on standby. As the competition in the first-tier markets becomes more and more intense, Chinese e-commerce enterprises are expanding their business to smaller cities and remote rural areas where the warehouses are dispersed and the infrastructure worse. As a result, automation has little to do.

Furthermore, fresh product online shopping in China has been booming recently. But the fresh product logistics are recognized as the most difficult. It is indeed a great challenge for all the e-commerce companies. Even robots as clever as Kiva can’t serve the fresh products warehouses of Amazon at this moment.

Theoretically, fresh product logistics, as challenging as they are, should have attracted large investment and innovation. But on the contrary, since the locations and sizes of the fresh product suppliers are very diverse and decentralized in China, what you can see is a cluster of small companies and individual venders, for whom automation is not necessary or affordable.

In conclusion, automation has become the main trend of all industries; ambitious companies along the logistics and supply chain should try to adopt new technology and innovations. However, more economic and localized solutions still need time. “It is estimated that over 90% of the human movement (in a warehouse) is useless, we must find a way to make our workers walk less, for example, by using more of the electronic pallets,” said Zhou Tao.

But we should always keep in mind that the most important thing is that any technology adoption must match the industrial ecosystem. Automation can be a good tool rather than a goal. Nowadays, building a solid information infrastructure and having warehouse process standardization is more important for the e-commerce and logistics industry. Only with these foundations can the “Santa Claus” control the reindeer (Kiva) well.