Self-driving cars – also known as “autonomous vehicles” – are becoming an increasingly common sight on the streets in cities around the world.
Tesla, Audi and GM are among the list of car manufacturers who already offer vehicles with limited Level 2 or 3 automation capabilities, including the option of hands-free automated steering under certain conditions (see “Degrees of Automation” sidebar). But truly autonomous vehicles (Level 5) that can operate without human drivers, while relying on artificial intelligence and LiDAR (Light Detection and Ranging) technology to navigate through traffic, are still being tested and refined.
Though the current autonomous vehicle prototypes still need significant road testing and improvement, many people expect driverless vehicles to, eventually, revolutionize the automobile industry, spur new electronics market growth, while improving the environment by decreasing pollution, traffic congestion and energy use.
Before that enticing vision is realized, however, a new generation of wireless networks capable of high-bandwidth, highly reliable communications with self-driving vehicles need to be designed and widely deployed.
Massive data volumes
Today’s cars offer an array of increasingly powerful Internet, smartphone, entertainment and navigation functions, and as automakers continue to add new tech capabilities, vehicles are having to deal with far more data than ever before. But the volume of data they’re processing now pales in comparison to what self-driving vehicles will need to monitor traffic, avoid obstacles and safely navigate their way through busy city streets.
Industry forecasters expect the first generation of autonomous vehicles to generate and consume up to 4 terabytes of data per day – the data equivalent of streaming more than 1,600 90-minute Netflix videos. Such massive data volumes are well beyond what current wireless networks can handle.
The auto industry faces a daunting challenge in figuring out how to manage the unprecedented volumes of data that self-driving vehicles will generate, and consume. To help streamline moving data from the car to the cloud and back, there is a new technology solution that appears suitable to meet the challenge. It’s fifth-generation (5G) wireless, the latest successor to the 3G and 4G technology used in most of today’s existing wireless voice and data networks. And it’s arriving just in time to help usher in this era of self-driving vehicles.
The wireless industry’s standard-setting group, the 3rd Generation Partnership Project (3GPP), is still hashing out final details of the 5G standard, but most of the essential features, functions and system architecture needed to deploy commercial 5G chips, modems and network equipment have now been agreed upon, and the initial performance results look promising. The new 5G technology is expected to be up to 100 times faster than current 4G networks, with significantly less lag time for executing instructions and more capacity for multiple users, thanks in part to its use of multiple antennas and additional radio spectrum.
5G testing underway
At least a dozen network operators, including AT&T, Verizon and Vodaphone, are field testing 5G network technology, with plans to launch the first fully 5G-compliant mobile networks in 2019. Meanwhile, semiconductor and electronic systems developers are already introducing their first 5G products. Samsung and its auto electronics subsidiary Harman International, for instance, recently announced a “5G-ready” telematics system capable of communicating with both today’s 4G networks and future 5G networks.
While communications companies are rolling out 5G networks, the auto industry continues to focus on automating driving technology. Several car models, for instance, can now warn drivers of potential collisions and assist with emergency braking. Some have systems to help with parking or to prevent drivers from straying out of their lane, while others (including Audi, GM and Tesla) offer hands-free steering assistance.
Industry research firm, IHS Markit, predicts that although ride hailing services will fuel the initial deployment of self-driving vehicles, individual buyers will join in after a few years, pushing worldwide autonomous auto sales to more than 33 million units a year in 2040.
It’s too soon to tell whether self-driving vehicles will live up to these optimistic projections, and how popular they will actually be with consumers. But clearly, while expectations are running high, the success of this promising market will depend on the availability of 5G wireless networks as well as the in-car autonomous driving technology.
# # #
Learn more on our landing page:
In 2014, the international Society of Automotive Engineers (SAE) defined the following six levels of automation that distinguish the varying capabilities of on-road, self-driving vehicles. The National Highway Traffic Safety Administration (NHTSA) adopted the SAE standard in place of its own classification system in 2016.
Level 0: No automation. Requires a full-time human driver. It may have regular cruise control, but not the ability to monitor and match the cruising speed of other nearby vehicles.
Level 1: Driver assistance. May include adaptive cruise control, as well as lane-keeping and parking assistance. But the driver needs to maintain full situational awareness and control of the vehicle.
Level 2: Partial automation. Can match the speed of surrounding traffic and follow curves in the road during normal highway driving. But drivers must stay alert and ready to resume control.
Level 3: Conditional Automation. Actively monitors the driving environment, changes lanes and navigates through traffic jams under certain conditions. Drivers still must intervene when requested.
Level 4: High automation. Handles most normal driving tasks, but may request occasional human intervention. If the driver does not respond appropriately, it may continue driving or decide to park itself.
Level 5: Full automation. Can drive itself anytime, anywhere, under any conditions, without human intervention. Steering wheel and other human driving controls are optional, and seats may face backward.
Tags: automotive, autonomous driving, Samsung, Samsung Semiconductor