To Make Cars Safer, Connect Them to Everything

V2X technology can reduce traffic and deaths even as autonomous vehicles stall

3 min read

overhead view of a traffic intersection with lines connecting vehicles

This article is part of our exclusive IEEE Journal Watch series in partnership with IEEE Xplore.

Between Elon Musk’s broken promises of “Full Self-Driving“ and California’s ban on Cruise’s driverless taxis, initial public enthusiasm for autonomous cars has given way to rampant skepticism and regulatory scrutiny.

Yet automakers, tech giants, and researchers continue to work the problem, even as Alphabet’s Waymo lays off employees, General Motors’ Cruise sheds billions of dollars, and Ford and Volkswagen’s Argo AI unit shut down entirely. Continued hope comes from a belief that autonomous vehicles and ones connected to communications networks could help to meet global goals to reduce emissions and limit injuries and deaths in cars. The role of sustainable mobility looms large over the European Green Deal, whose 2050 deadline dovetails with a parallel “Vision Zero” target in the United States to eliminate roadway fatalities.

The new field of vehicle-to-everything (V2X) connectivity, in which individual cars would communicate wirelessly with one another as well as traffic signs and other infrastructure, could help reach those targets. A paper published 11 March in the IEEE Open Journal of Vehicular Technology argues that V2X—underpinned by 5G and 6G networks—can “pave the way toward safe, affordable, accessible and sustainable transport systems, as well as improve road safety.” As the paper notes, if V2X could dynamically coordinate traffic, the result would be less congestion, freer-flowing transportation, safer streets, and reduced emissions.

In 2021, more than 7,300 pedestrians were killed in the United States.

The researchers in Spain and Sweden who authored the paper focused on future smart cities with high population density. They identify three current trends leading toward sustainable V2X communications: Climate-neutral infrastructure and vehicle-user equipment; “cloudification” and edge computing; and big data and AI.

The researchers envision a model with roadside units linked to both cellular and cloud-computing networks. These units would be able to communicate with smart traffic signals, road sensors, units on board individual vehicles, and edge and cloud nodes. There’s also a need to connect pedestrians and bicyclists so that vehicles and vehicle infrastructure know they’re there—either through passive detection or by actively signaling the network with onboard Internet of Things devices.

Integrating pedestrians into a V2X system could help stem a troubling rise in pedestrian deaths in the U.S. since about 2009, especially at night. In 2021, more than 7,300 pedestrians were killed in the United States, and 75 percent of those deaths occurred between sunset and sunrise.

The authors note the potential for an Internet of Vehicles—a network of connected autonomous vehicles, in other words—that could boost communications efficiency by optimizing transmission rates, channels, and times. An Internet of Vehicles could create dynamic 3D maps of traffic and pedestrian flow, communicate object locations between cars to avoid collisions, and monitor air pollution and vital signs of drivers or passengers in vehicles.

One challenge of connecting every car to a wireless network is the massive amounts of generated data. In response, the researchers outline the potential for 6G network trends—including virtualization, moving network functions to the cloud, and edge computing—to handle the load.

There’s also a place for AI in V2X systems. AI can enable real-world and real-time fuel and emissions monitoring, as well as vehicle platooning, in which vehicles travel together in autonomous, tightly spaced chains to save fuel and free up road space. It could also help with speed harmonization to gradually lower the speeds before congested areas and therefore eliminate stop-and-go traffic.

With V2X and autonomous cars both in their infancy, Sandra Roger, one of the paper’s authors, an IEEE Senior Member, and associate professor at Valencia Polytechnic University, acknowledged enormous challenges ahead, including those that technology alone can’t address. For one thing, data-privacy advocates and fans of old-school human driving will likely shudder at the idea of centralized monitoring and control over streets and roads, despite potential upsides.

“Use cases such as speed harmonization would involve convincing drivers to leave some or all control over vehicle speed to automated systems, and this may face resistance due to trust issues and perceived loss of control,” Roger said in an email to IEEE Spectrum.

Roger identified other challenges, such as legal liability in accidents, cybersecurity and privacy concerns, and regulatory agreement on vehicle-to-vehicle (V2V) protocols and safety standards. She is also concerned about standardization and interoperability among automakers and other manufacturers.

Infrastructure upgrades are another heavy lift. “Upgrading traffic signal systems to accommodate connected vehicles can be costly and disruptive, especially in urban areas with heavy traffic,” Roger said.

Still, the authors recognize that V2X likely will not be in existence everywhere. They conclude: “Addressing sustainable mobility in rural or suburban areas will need a different solution,” since V2X deployment seems unlikely due to those cost-benefit trade-offs.

From Your Site Articles
Related Articles Around the Web
The Conversation (1)
Andreas Kwoczek
Andreas Kwoczek20 May, 2024

As long as these connected vehicles use the IEEE802.11p standard instead of operator depending 4G, 5G or 6G technology. There will be no business case for the operators or the necessary entity to manage the traffic between them to reach every vehicle. We could have saved thousands from fatal acvidents, if it wasn't for the lobbyist of the operators still hoping to introduce 3GPP technology to make money. As for pedestrians we have to solve the positioning/localization down to 10-20 cm before can include them in any communication. Otherwise we end up in false collision warnings.