Advanced Computing in the Age of AI | Monday, August 8, 2022

Autonomous Trucks Demo Uses Nvidia AI Platform 


A test fleet of autonomous delivery trucks scheduled for deployment next year will be outfitted with self-driving system based on Nvidia's AI processing technology for autonomous vehicles.

The AI car computer called Drive PX will be combined with German automotive supplier ZF's self-driving platform in a fleet test planned by package delivery and logistics vendor Deutsche Post DHL Group (DPDHL). The goal is solving one of the biggest challenges for automating package delivery: getting deliveries the "last mile" between a central location to their final destination.

The 2018 demonstration will use the package delivery company's (ETR: DPW) fleet of 3,400 electric delivery vehicles outfitted with cameras, radar and lidar (light detection and ranging). Sensor data is fed into an AI platform based on Drive PX. The partners claimed the combination can leverage AI and deep learning to allow autonomous vehicles to understand its surroundings, plot and drive along a safe route, and then park itself at the delivery point.

"The development of autonomous delivery vehicles demonstrates how AI and deep learning are also reshaping the commercial transportation industry," Nvidia CEO Jensen Huang noted in a statement announcing the partnership during a company event this week in Munich, Germany.

"As online shopping continues to explode, and the shortage of truck drivers becomes more dire, AI-enabled vehicles will be key to providing last-mile delivery services," Huang added.

DPDHL said it has configured its datacenter with Nvidia's (NASDAQ: NVDA) widely used DGX-1 AI platform used to train neural networks. Deep learning models will be run on the Drive PX platform. Prototype delivery vehicles are equipped with six cameras, two lidars and radar that will feed location and other data into the Nvidia system.

The autonomous delivery scheme "follows our 'see - think - act' approach," explained ZF CEO Stefan Sommer. "In supply logistics and on the last mile where autonomous driving has tremendous benefits, goods can be delivered independent of the time of the day and delivery staff, with minimal noise and emissions, thus significantly reducing traffic congestion in city centers."

The Nvidia platform is designed to scale from a single to multiple processors and discrete GPUs capable of driving autonomous vehicles. Drive PX combines deep learning, sensor fusion and machine vision with the ability to figure out location using onboard maps to plot a safe course.

The sensor fusion capability includes processing lidar and radar data along with algorithms that spot obstacles along a route. Nvidia said it also uses deep neural networks to detect and classify stationary and moving objects as a way to increase the accuracy of fused sensor data. Those networks can be trained in datacenters prior to deployment in autonomous vehicles.

The resulting model runs in real time on the Drive PX installed in a vehicle.

Separately, Nvidia unveiled what it claims is the first AI computer designed to drive fully autonomous "robotaxis."

The new system, codenamed Pegasus, extends its Drive PX platform to handle driverless vehicles. Pegasus delivers more than 320 trillion operations per second, the company said this week. The system is designed to run in fully autonomous vehicles without steering wheels, pedals or mirrors.

--Editor's note: This story has been updated.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

Add a Comment