The countdown has started for widespread use of driverless taxis in Chinese cities with new developments in autonomous control
After gaining driving permits for driverless testing on public roads in several major cities across China last year, US and China based autonomous driving technology company, Pony.ai has continued to make inroads on the core technology at the heart of their vehicles to create an extensive testing and delivery programme for 2022.
The latest permit the company received in December last year was in the Chinese city of Shenzhen, the home of Pony.ai’s research & development centre. The company’s autonomous vehicle services are now being tested across approximately 300 square miles throughout China.
With the addition of Shenzhen, Pony.ai has now launched autonomous driving tests or service operations in a total of seven cities in China and the USA. It is also one of the first two companies to commence autonomous driving tests in all four of China’s Tier-1 cities.
Since 2017, Pony.ai has accumulated millions of autonomous real-world road test mileage, with extensive hardware and software development experience in Beijing, Shanghai, Guangzhou, and California. Pony.ai’s autonomous vehicles have thrived in complex driving conditions, including heavy urban traffic, complex road and tunnel layouts, extreme weather, and thousands of challenging scenarios, all of which underscores the company’s decision to move platforms to the most state-of-the-art equipment with enhanced technological capabilities.
Pony.ai believes that testing across such diverse territories at a meaningful scale strengthens the company’s ability to provide safe and reliable autonomous driving globally.
Automotive Grade Silicon
The bullish outlook for 2022, including the forecast of commercial service deliveries within the year, is based on the move Pony.ai has made from what it refers to as industrial-grade processing hardware to processors that have a higher automotive-grade capability.
The complete autonomous computing unit that Pony.ai is using is built on the NVIDIA DRIVE Orin system-on-chip (SoC) hardware and it is expected to rapidly accelerate the roll-out of Pony.ai’s Level 4 autonomy capabilities and is expected to be mass-production ready with the start of production slated for the end of 2022.
NVIDIA DRIVE Orin achieves 254 TOPS (trillion operations per second) of performance and includes comprehensive CUDA and NVIDIA deep learning accelerator (NVDLA) toolchain support.
Pony.ai’s autonomous computing unit features low latency, high performance and high reliability. The company is one of the first in the AV industry to create a product portfolio featuring multiple configurations with one or more DRIVE Orin processors and automotive-grade NVIDIA Ampere architecture GPUs. This enables scalable deployment across self-driving trucks and robotaxis, and accelerates Pony.ai’s future of a robust, mass-production platform for AVs.
According to James Peng, Co-Founder and CEO of Pony.ai, by making use of the world-class NVIDIA DRIVE Orin SoC, the company can demonstrate its design and industrialisation capabilities plus the ability to develop and deliver a powerful mass-production platform at an unprecedented scale.
“This will help us realise our vision of bringing safe autonomous mobility systems to all,” he says.
Gary Hicok, Senior Vice President of Engineering at NVIDIA, added, “As a close collaborator with Pony.ai since their inception, we applaud their vision and execution with the development of their AV system architecture to bring safe, Level 4 capabilities to market.”
This relationship shared between Pony.ai and NVIDIA dates back to 2017 when the company first adopted the NVIDIA DRIVE platform. Last May, Pony.ai began co-developing DRIVE Orin-based systems with NVIDIA, making use of the high-performance SoC to accelerate the pace of hardware development.
Already starting the next generation of its autonomous drive project, Pony.ai is now testing its Robotaxi vehicle, which employs advanced sensor systems, the NVIDIA DRIVE computing platform and a styling and design package, which suits level 4 automotive-grade fleets for mass production.
The Robotaxi uses the Toyota S-AM, a seven-seat hybrid electric platform for autonomous mobility and is starting its road testing regime in China this year with an expectation that it will be deployed in public operations during the first half of 2023.
According to Peng, Pony.ai was founded with on the basis of developing a route to the mass deployment of Autonomous Driving technology and has so far accumulated more than 10 million kilometers of autonomous mileage worldwide on open roads. It will continue to invest in R&D and global robotaxi operations throughout 2022.
“The introduction of the world-class Toyota S-AM into the PonyPilot+ fleet represents another major achievement of the deep strategic cooperation between Pony.ai and Toyota. Furthermore, Pony.ai’s new generation of autonomous driving systems powered by NVIDIA and designed for L4 automotive-grade mass production, as showcased on the S-AM, will further prove the safety and technical reliability – and promote the large-scale application – of AD technology,” says Peng.
On the basis of its work with Toyota on developing the self-driving capabilities for the S-AM China model and through extensive engineering and testing, Pony.ai believes this highly customized S-AM with its dual redundancy system, will provide better functionality and control performance for Level 4 AD development. In addition, as an MPV model, the S-AM has a flexible riding configuration to meet numerous family and professional usage scenarios.
Cutting edge technology
To prepare the Robotaxi for its full deployment, it has been fitted with the latest sensor and computing technology available in the automotive industry. The new autonomous driving system is a major breakthrough in software, computing capability and redundancy, according to Pony.ai’s co-founder and CTO Tiancheng Lou.
The sensor system, containing a total of 23 sensors, includes four solid-state LiDARs on the roof, covering a 360° field of view; three near-range LiDARs on the body of the vehicle, covering the blind spots of the roof LiDARs; four millimeter-wave radars located at the corners of the roof; one long-range forward-facing millimeter-wave radar, and 11 cameras deployed around the roof and body of the vehicle (in a combination of wide angle, super wide angle, middle and long-range, and traffic light detection cameras).
The central mechanical LiDAR has been replaced with solid-state LiDARs. The self-developed traffic light camera has a resolution rate 1.5 times that of the previous generation. Coupled with its in-house Sensor Fusion technology, Pony.ai will significantly reduce the cost associated with the detection systems by utilising mass-produced, automotive-grade sensors.
Using the NVIDIA DRIVE Orin processor, Pony.ai’s autonomous computing unit features low latency, high performance and high reliability. The company is one of the first in the autonomous vehicle industry to create a product portfolio featuring multiple configurations with one or more DRIVE Orin processors and automotive-grade NVIDIA Ampere architecture GPUs. This enables scalable deployment across self-driving trucks and robotaxis, and accelerates Pony.ai’s future of a robust, mass-production platform for autonomous vehicles.
Compared with the previous generation computing platform, the new generation is expected to have greater than a 30% increase in computing power, at least 30% less weight, and more than a 30% reduction in cost.
The new generation system has complete redundancy in place, which maximises safety and enables the vehicle to pull over or stop safely in case of an emergency.
A Sleeker Design for Robotaxi
The technology set, including the sensors, has allowed Pony.ai to concentrate on the aesthetics of the vehicle. The previously obtrusive looking roof design has been re-shaped based on the lower profile of the LiDAR sensor housings to give the car a better overall appearance.
The concept design of the S-AM model and the more integrated and aesthetically pleasing sensor suite has also made the vehicle more effective for mass production.
Design features also include rooftop signalling units which have a horizontal lighting unit on the front and three vertically placed lighting stripes on the rear, to give the vehicle an elegant futuristic touch. By using different colours and lighting combinations, the rooftop signalling units can provide external communication and interaction, demonstrating the robotaxi’s status and intentions.