Qualcomm’s Snapdragon Ride Platform drives innovation en route to fully autonomous vehicles
Rapid advancements in technology and a collective love of sci-fi have Americans all revved up for a future driven by fully autonomous vehicles. It’s an exciting prospect: highways zipping with cutting-edge cars that are navigated by artificial intelligence while passengers kick back behind the wheel, and after half a century of dreaming and decades of development, autonomous driving (AD)—to the extent of full self-driving (FSD)—is closer than ever to being a reality.
Still, the journey to the fully self-driving car will not be a non-stop ride. Already, it has been rocky for early aspirants, who have faced significant roadblocks along the way, including vehicle testing and regulations that differ from country to country. Educating consumers and gaining their trust has been another challenge during the deployment and adoption process. In its recent American Experiences Survey, Consumer Reports found that only 37 percent of Americans correctly answered that fully self-driving cars are not yet available for purchase in the U.S. today. Nearly 29 percent of Americans said that self-driving cars are available for purchase, and 34 percent were unsure. These findings indicate uncertainty about what today’s cars are capable of.
“During the earlier days on the journey to autonomy, there were many claims and discussions in the industry about how we were getting prepared for automated driving, and many in the automotive and mobility industries were incredibly eager to explore it,” says Anshuman Saxena, vice president, product management and head of ADAS/AD products at Qualcomm Technologies, Inc. “However, they soon after realized that the complexity of a self-driving automated system is pretty high and dependent on many factors.”
For self-driving cars to be viable, the industry must first further develop and master advanced driver assistance systems (ADAS), which are already beginning to take to the road. For the most part, automakers that produce passenger cars will zero in on ADAS development, with an emphasis on improving the safety and comfort of vehicles, which the industry considers as lower levels on a five-point scale of autonomous capability.
“At this time, we see automakers continuing to focus on improving active safety and convenience as we work together to push ADAS and AD forward—they’re not necessarily making the claims for a level four or five of autonomy,” Saxena says. “As we look ahead, in addition to moving forward with passenger vehicles, there will be continued technology advancements with robo-taxis, robo-buses, long-haul trucks and driverless delivery, with an early deployment for the fully autonomous driving systems towards the later part of this decade.”
Automakers, hardware manufacturers, and software experts are alternatively competitive and collaborative in their development of the technology that powers ADAS. Uniting nearly all players is Qualcomm Technologies’ Snapdragon Ride Platform, a scalable and open ADAS and autonomous driving solution that offers unprecedented capabilities and flexibility to automakers and developers.
Qualcomm Technologies is continuing to build global momentum in advanced driver assistance and the autonomous driving segment with its Snapdragon Ride Platform. The platform is driving innovation for top-tier automakers, in whichever direction they seek to advance, providing equal access to safe and more comfortable driving experiences to vehicles across all segments. When it first launched for automakers in 2020, the platform consisted of three main pillars: Snapdragon Ride Safety system-on-chips (SoCs), the Safety Accelerator, and the Autonomous Stack.
The acquisition and integration of technology from Arriver has since facilitated the creation of the Snapdragon Ride Vision System as well as the Snapdragon Ride Flex SoC – the automotive industry’s first scalable family of SoCs to simultaneously support mixed critical systems — digital cockpit and advanced driver assistance systems. The Snapdragon Ride Flex SoC is pre-integrated with the Snapdragon Ride Vision stack, which enables highly scalable and safe driver assistance and automated driving experience using sensors like front camera to meet regulatory requirements. Additionally, it has multi-modal sensors for enhanced perception that creates an environmental model around the vehicle feeding into vehicle control algorithms.
Paired with the Snapdragon Ride Software Development Kit — a toolset that helps automotive OEMs and developers build safety-certified applications for perception and drive policy — the Snapdragon Ride Platform now allows automakers and Tier-1 suppliers to develop powerful and efficient designs, as well as integrate their own proprietary map crowdsourcing, driver monitoring system (DMS), parking systems, cellular vehicle-to-everything (C-V2X) technologies and localization modules.
“We’re working on improving efficiency and comfort by easing those regular drives like daily commutes to work, while also improving safety,” Saxena explains. “From a vehicle maintaining control of itself to avoid accidents if a driver gets distracted or eventually even negotiating the traffic lights and driving its way through, these are some innovative auto safety features we’re developing within ADAS and AD.”
Snapdragon Ride is designed to meet the highest level of automotive safety while enabling a hardware architecture to support isolation, freedom from interference, and quality of service for specific ADAS functions.
One of chief challenges is the sheer collection and quick analysis of unprecedented amounts of data required to create that sort of intuitive driver system. The ADAS system needs to be prepared for all possibilities and variables, and leans on an enormous number of cameras and machine learning to make sense of it all.
“You have to understand the information cameras are giving and decode it to identify corner cases and driving patterns of our driver versus other drivers,” Saxena says. “The system becomes retrained based on new patterns and corner cases detected. The efficiency of collecting the relevant data and having a mechanism of continuous learning to improve the systems is the key for faster deployment of higher levels of ADAS.”
The Snapdragon Ride Platform’s scalability serves to immediately begin the democratization of the ADAS transformation. Qualcomm Technologies has already been tapped by the world’s leading auto manufacturers who have sought the flexibility of the platform to meet their specific needs. For example, Qualcomm Technologies has become the go-to partner for developing high-performance hardware solutions for hands-free driver assist systems. The company also collaborates with automakers to produce software stacks that address the full range of ADAS challenges up to level three autonomy.
The technology is an ideal central-compute solution that supports next-generation software-defined vehicle (SDV) solutions for OEMs and across the SDV ecosystem. OEMs can tap into new software services models while providing their customers with an always-improving user experience.
As a scalable platform, Snapdragon Ride allows automakers to develop their own applications and capabilities, as well. Instead of putting the fate of automated driver assist systems in the hands of just a few automakers, Snapdragon Ride provides the opportunity to revolutionize the technology and auto industry to everyone.