AINEWS
An IEEE Spectrum column argues the open cellular standard behind your phone could be repurposed to let cars share sensors, though no such test network exists yet.
Two cars roll toward the same blind intersection on a four-lane road. One is a delivery van, the other a sedan. Each is a data center on wheels, packed with cameras, radar, and lidar. Neither can see the other: the delivery van's roof racks block the sedan's sensors, and a row of parked SUVs hides the pedestrian stepping out between them.
That deafness is what V2X, or vehicle-to-everything communication, has been trying to fix for more than a decade. A fresh first-person analysis in IEEE Spectrum from a V2X researcher lays out an argument for what the fix should look like, and why a decade of purpose-built protocols have kept failing at it.
Most self-driving stacks today lean on the "data center on wheels" model. Each car carries enough compute and sensors to model the world on its own. The proposed alternative is a community antenna on wheels: instead of every car seeing the road alone, nearby cars share what their sensors and vantage points can see, so a sedan that cannot spot a freight truck around a curve gets that information from the truck itself, or from a second car two blocks back that has a clear line of sight.
The proposed plumbing is O-RAN, the open, programmable architecture that some 4G and 5G cellular networks already run on. O-RAN was built around the assumption that cell towers are fixed, but it can be taught to treat cars and trucks as mobile towers with relatively little modification. Once a car is a tower, the radios, scheduling software, and the broader cell ecosystem are already in place to handle handovers, congestion, and security. Because the standard is open, any carmaker, city, or vendor can plug into the same ecosystem without paying a single proprietary gear supplier for the radio stack. A working version could let one car warn another about a pedestrian it cannot see, or fill in a sensor blind spot from a truck's line of sight.
The argument is to stop inventing new V2X protocols from scratch and borrow one that already works at city scale. The author frames the past decade of purpose-built vehicle networks, including DSRC (the dedicated short-range radio standard) and C-V2X sidelink (the cellular peer-to-peer variant), as a string of "new networking protocols built from scratch" that kept running into the same three walls: a chicken-and-egg problem where no automaker would deploy until the others did, a standards war over which protocol to pick, and a real-signal problem where simulators assumed perfect radio conditions that moving, obstructed, high-density traffic never has. In a real city, signals bounce off buildings, get blocked by trucks, and arrive at the wrong moment because cars are moving at 30 mph. Most V2X academic studies have used models that assume those conditions are clean, which is part of why the field's optimistic simulation results have not translated to deployment.
The author has a precedent for the borrowing move. Wi-Fi, the most successful wireless standard of the last thirty years, was repurposed from a 1990s Australian radio astronomy project that was trying to detect evaporating black holes. Networking history is full of solutions lifted from unrelated problems; this argument asks the cellular industry to do the same.
The intellectual scaffolding for an O-RAN-plus-V2X design is already in the published literature. An arXiv preprint from 2023, revised in 2024 by Polese and colleagues, framed in a longer writeup as making the case that vehicular communications integrated with the radio access network is "a breakthrough application for the 6th generation (6G) cellular systems," argues that traditional radio access networks lack the flexibility for V2X's strict latency and reliability requirements.
The author states it: "There is, to be clear, no O-RAN V2X test network operating in the world. Not yet." The argument is for a direction, not a deployment. The arXiv work is a preprint, not a peer-reviewed field trial, and the IEEE Spectrum piece is one researcher's advocacy column, not industry consensus.
What to watch: a real test network. A city-scale pilot with moving vehicles acting as mobile O-RAN nodes, even at a single intersection, would move the conversation from borrowed theory to evidence. None is operating yet.