Menu Close and Nvidia’s smart city nears completion

Each smart delivery robot is equipped with two Nvidia Jetson AGX Xavier platforms, with the largest capable of delivering more than 500 orders


Graphics processing units (GPU) inventor Nvidia and Chinese retail giant are building the first smart city with the world’s first Level-4 smart delivery vehicle application in China.

Nvidia and have been working together since 2017. Making intelligent robots based on the Nvidia Jetson platform, JD brings AI to the field of logistics and supply chain, and continues to expand its research in autonomous driving and smart delivery.

“JD’s smart delivery vehicle is the world’s first application of Level-4 autonomous driving technology on public roads without any human interaction,” Cindy Xu, Principal Architect of
JD Logistics’ Autonomous Driving division in Silicon Valley, told Nvidia’s online GPU Technology Conference 2020.

Immediately after the outbreak of COVID-19 in early 2020, JD Logistics used its Level-4 autonomous delivery vehicle in Wuhan to ensure contactless delivery from delivery stations to hospitals and residential compounds. Within 107 days, the robots had travelled more than 6,800 kilometres and delivered more than 13,000 packages.

In October, JD Logistics announced it would help Changshu, Jiangsu Province, a top five county-level city by GDP in China, build a smart city based on Level-4 smart delivery vehicles.

JD, JDL, LD,com
A JDL employee loads the autonomous robot that delivers orders in Changshu, China.

More than one in five delivery stations in Changshu has been equipped with smart delivery vehicles. JD Logistics plans to put 100 smart delivery robots into operation in the city by the
end of this year.

Each vehicle is equipped with two Nvidia Jetson AGX Xavier platforms, the largest of which has a cargo space of two cubic metres and can deliver more than 500 orders at a time.

The latest smart delivery robot can drive itself on open roads without human control and plans its own route based on the address on the packages. It will send messages to
recipients with a verification code so they can pick up their packages.

With the growth of online shopping platforms, delivery volume has skyrocketed over the past few years. According to the State Post Bureau of China, nearly half of the couriers in
China work 10 to 12 hours a day and the country delivered more than 60 billion packages in the first 10 months of 2020.

The project aims to alleviate couriers’ workload, with JD launching a delivery robot “adoption plan.”

In Changshu, couriers can apply to adopt the delivery robots to assist their work. In such cases, couriers can focus on parcel collection and more customised services.

“The core of achieving this level of research and application is the deep collaboration between JD and Nvidia. Based on Nvidia’s Jetson AGX Xavier platform, we have further developed our autonomous driving system and optimised the computing unit,” Xu told the Nvidia GPU Technology Conference.

“The collaboration with Nvidia not only secured the Level-4 autonomous driving demand from a computing perspective, but also helped lower the power consumption to 10 per cent of the industry average.”

During this year’s Double 11 (November 11) shopping festival, couriers who adopted delivery robots showed a 50–100 per cent increase in the number of orders they delivered, <em>Pandaily</em> reported.

China is the global leader in the autonomous delivery sector.

Leave a Reply