While Alibaba just announced that its fleet of 200 autonomous robots delivered more than 1 million parcels in China within a year of launch, the e-commerce giant has even bigger goals for what it calls the “Xiaomanlv.” By 2024, Alibaba plans on operating 10,000 of these last-mile delivery robots in a quest to handle an average of 1 million packages a day.
Collectively, the robots have delivered packages to more than 200,000 consumers in 52 cities across 22 Chinese provinces. Each robot can carry about 50 packages at a time and cover 100 kilometers on a single charge, delivering nearly 500 boxes per day.
According to Alibaba’s global research arm, Damo Academy, the robots leverage a cloud-based intelligent simulation testing platform, which can create up to 10,000 virtual scenarios, accounting for extreme weather and poor visibility at night. Xiaomanlv can operate without human intervention 99.9999 percent of the time, the research arm claims, despite needing to identify more than 40 million obstacles during a typical day’s work.
Various algorithms enable the robots to detect these small obstacles, navigate precisely and identify the intentions of pedestrians and vehicles within fractions of a second.
“One important breakthrough is our ability to use advanced algorithms to achieve a low-cost mass deployment of our self-driving vehicles across communities and campuses,” said Gang Wang, head of the autonomous driving lab at Alibaba Damo Academy. “Through our proprietary multi-sensor fusion solution and cutting-edge machine learning platform AutoDrive, we can achieve the ‘L4’ level of autonomous driving without leaning on expensive, high-definition sensors to navigate. This helps us significantly cut down our hardware costs. As a result, the overall per-unit cost of production and operation is just about one-third of the industry average.”
L4, or Level 4, refers to a scale of Level 1 to 5 in autonomous driving, with Level 4 meaning a high degree of automation.
Xiaomanlv, which loosely translates to “little competent donkey” in English, initially debuted on the Zhejiang University campus in preparation for the e-commerce giant’s 11.11 online shopping festival, known as Singles Day.
The same technical framework that powers the Xiaomanlv is now being used to design self-driving delivery trucks, in collaboration with Alibaba’s logistics arm Cainiao Network, Wang wrote in a blog post.
“These larger robots will be able to deliver goods at a faster speed and for longer distances. They are likely to appear on public roads within three years,” Wang said.
Alibaba has been testing autonomous delivery since 2015 to address last-mile challenges, but Wang said the 1 million autonomous deliveries milestone “is just the beginning.”
The overall costs of producing and operating Xiaomanlv is one-third of the industry average, according to Wang. The price competitiveness comes from lower hardware costs, as the robots operate based on trained algorithms instead of the high-definition sensors used by most self-driving cars. Alibaba uses a proprietary multi-sensor fusion solution and AutoDrive, its machine-learning platform, to power Xiaomanlv.
Damo Academy is also developing other products equipped with autonomous driving capabilities, including inspection robots in electricity power plants, which are expected to be used in the near future.
Alibaba is looking to further expand the last-mile capabilities beyond China as well, with the Cainiao Network recently launching a locker network in Spain and France.
While Cainiao’s network currently spans 170 lockers in Madrid, Barcelona and Paris, the firm plans on expanding the program to 2,000 across the two countries by March next year.
The contactless lockers enable shoppers who buy from platforms like AliExpress to pick up their parcels at flexible times, while reducing the carbon footprint of last mile services, according to Alibaba.
Alibaba and rival e-commerce player JD.com have both been making significant investments in robotics, and this only ramped up when the Covid-19 pandemic forced Chinese consumers to pivot further to online shopping when they were quarantined in their homes. JD.com’s biggest splash came in February last year, when the company deployed a fleet of unmanned robotics in the Chinese city of Wuhan, the pandemic’s original epicenter. In a span a little over a week, from Jan. 24-Feb. 2, JD.com’s automated warehouses saw daily orders increase from 600,000 to 1 million.