Last time we talked about road paint, Herman mentioned that we're already moving into an era where physical signs become legacy support. Daniel read that and had questions — a whole stack of them. He wants to know how this works technically under the hood. We've got stop signs, lane markers, painted road surfaces, signs on poles — and from what he understood, all of these are already specced for fully digital replacements where the information gets communicated directly to smart cars. He's asking what NASN specification has been put down, what's being developed, when we might see the first rollout — and the big one: will physical signs be completely replaced, leaving an all-digital system in place?
So let's start with the thing that sounds the most sci-fi and work backward. How does a road sign that isn't there tell a car what to do?
Because that's the part that sounds like someone's been reading too much speculative fiction.
It does, but the spec exists and it's been running in real deployments for about five years now. The core thing to understand — and this is where most coverage gets it wrong — is that digital road signage doesn't mean a screen on a pole. It's not a Kindle taped to a stop sign post.
Thank goodness, because that would be the ugliest infrastructure decision since the concrete brutalist parking garage.
What it actually means is that the road itself — or more precisely, a box mounted near the road — broadcasts machine-readable messages directly to vehicles over short-range radio. Speed limits, stop conditions, lane configurations, what the traffic light is about to do. All of it arrives as data packets, not paint and not LEDs.
So the sign is invisible to humans. A driver in an old car sees nothing.
And that's the central design problem the standards bodies have been wrestling with for two decades. If the signs are invisible, how do legacy cars, pedestrians, and cyclists know what to do? The answer — and we'll come back to this — is that physical signs remain mandatory. The digital layer is a supplement, not a replacement. At least for now.
Let's get into the radio side. What's the band, what's the protocol?
Five point nine gigahertz. The FCC allocated it for intelligent transportation systems back in 1999, then partially reallocated a chunk of it in 2020 — which was its own drama, the cable industry wanted it for Wi-Fi. What's left is still enough for vehicle-to-everything communication. Two competing protocols sit on that band. One is DSRC — Dedicated Short-Range Communications — which is based on the 802.11p standard, essentially Wi-Fi tweaked for fast-moving vehicles. The other is C-V2X, Cellular Vehicle-to-Everything, which uses LTE and eventually 5G New Radio sidelink. They both do the same job, they just do it differently, and which one wins depends on which country and which automaker you ask.
And the U.S. hasn't picked a winner.
The FCC has basically said both can operate in the band. The market's supposed to sort it out. Europe is leaning C-V2X. China's gone all-in on C-V2X. The U.S. is a mess — Ford committed to C-V2X in 2022, a lot of the DOT pilot projects used DSRC, and the 2024 national deployment plan is protocol-agnostic. It just says "V2X" and lets the implementers choose.
So the car's got a radio listening on five point nine gigahertz. What's it listening for?
This is where SAE International comes in. They maintain the standard data dictionary — SAE J2735 — that defines the message types. And there are three that matter for what Daniel's asking about. First, SPaT — Signal Phase and Timing. That's the message that tells a car exactly when the traffic light ahead is going to change, down to the millisecond. Not "green for a while longer" — actual countdown data.
Which is wild. The intersection is telling you its internal state.
Right. The traffic controller already knows when it's going to cycle — SPaT just broadcasts that timetable. Second message type is MAP. That's the geometry of the intersection. Lane widths, where the stop bars are painted, which movements are permitted from which lanes, the curvature of the turn radius. It's essentially a digital survey of the physical intersection. The car receives MAP data and can place itself within the intersection with centimeter-level accuracy, even if GPS is degraded.
So the car knows the shape of the road before it sees it.
Before its cameras or lidar ever get a look. And the third message type is the one Daniel's really asking about — RSI, Road Sign Information. That's the digital equivalent of a stop sign, a speed limit sign, a no-turn-on-red placard, a yield sign, a school zone warning. The roadside unit — the RSU, which is that box on a pole I mentioned — broadcasts an RSI message containing the sign type, pulled from a standardized code in the J2735 dictionary, its GPS position, and its validity period. The vehicle's onboard unit receives it, checks that the position is relevant to where the car actually is, validates the security certificate, and then either presents it to the driver on the dashboard or feeds it directly into automated driving decisions.
So the car gets a data packet that says "stop sign at these coordinates, effective immediately, signed by a trusted authority."
That's it. And the security layer is the part that actually makes the system viable. Every message is signed using IEEE 1609.2 certificates. The vehicle has a certificate authority chain baked in — it trusts certain root CAs, and those CAs sign the certificates that the RSUs use. If a malicious actor sets up a transmitter and starts broadcasting fake stop signs at an intersection, the vehicle's onboard unit checks the signature, finds it invalid, and rejects the message.
Which is great until the certificate expires.
That's exactly where it falls apart in practice. Certificates have lifetimes — typically a week to a year depending on the implementation. The RSU needs to periodically connect to a backend system to get fresh certificates. If the network connection lapses, or the cellular modem contract doesn't get renewed, or the backend server goes down for maintenance, the RSU can't sign messages with a valid certificate. The vehicle rejects them. The digital sign goes dark. And if you're relying on that digital sign for a safety-critical decision —
You're now relying on nothing.
Which is why the physical sign is still there. The spec is designed around the assumption that the digital layer can fail.
Let me ask the question Daniel asked directly. NASN specification. What is it, where is it?
Here's the honest answer: there isn't one. No NASN specification exists in any of the standards documentation, any of the DOT deployment plans, or any of the international harmonization efforts. I dug through the SAE catalog, the IEEE working groups, the ISO standards, the European C-Roads deliverables — nothing called NASN.
So where did Daniel get the term?
Best guess — and this is speculation — he may have conflated a specific project name or a vendor's internal designation with the broader standards ecosystem. There's a company called NaaS Technology, but they do electric vehicle charging infrastructure in China. There's the National Association of State Nine-One-One Administrators — NASNA — which deals with emergency communications, not road signs. There was a European project called "NAVI" that dealt with in-vehicle signage. None of them are NASN.
So the real work is distributed across a family of standards.
SAE J2735 for the message definitions. SAE J2945 for the performance requirements — minimum broadcast intervals, maximum latency, how often an RSU has to repeat its messages. IEEE 1609.2 for security. IEEE 1609.3 for network services. The whole thing sits under the umbrella of the U.S. DOT's Connected Vehicle program, and internationally under the ITS-G5 and C-V2X frameworks. There's no single spec with a catchy name. It's a stack.
Which is somehow less satisfying.
It's engineering. It's never satisfying.
Walk me through what happens when a car actually receives an RSI message. Paint me the moment.
Alright. You're driving down a road at forty miles an hour. Up ahead, there's a stop sign — a physical one, painted red, reflective sheeting, the whole thing. About three hundred meters before you reach it, the RSU mounted on the same pole broadcasts an RSI message. The message contains a data element called a SignType — that's a numeric code from the J2735 dictionary. Code one twenty-three might be "stop sign." Code one twenty-four might be "yield." The message also has a Position3D giving the sign's latitude, longitude, and elevation, and a MutcdCode that cross-references it to the Manual on Uniform Traffic Control Devices — the federal standard for what a stop sign looks like.
So the digital message points back to the physical standard.
Yes. The digital sign says "I am the digital twin of a physical sign that conforms to MUTCD section 2B.05." Your car's onboard unit receives the message, decrypts and validates the signature, checks that the sign's position is within your path of travel, and then does one of two things. If you're driving manually, it lights up a stop sign icon on your dashboard display and maybe sounds a chime. If the car is in some level of automated driving, it feeds the stop condition into the path planner, which begins decelerating to stop at the stop bar position defined in the MAP message for that intersection.
So the car doesn't need to see the physical sign at all. It could be fogged over, covered in graffiti, bent sideways by a snowplow —
And the car still stops. Because it got the message three hundred meters back, before its cameras would have had line of sight. That's the promise. Whether it works that reliably in practice is another question, and we'll get to the rollout reality. But that's the technical architecture.
So let's get to the rollout reality. When does this actually start showing up?
We're in what the DOT calls "early deployment." The V2X Deployment Plan released in 2024 sets phased targets — twenty percent of the National Highway System equipped by 2028, fifty percent by 2030, eighty-five percent by 2036. And I want to be precise about what "equipped" means here. It means RSUs broadcasting SPaT and MAP and RSI messages at intersections and along highway segments. It doesn't mean every stop sign in the country has a digital twin. It means the major corridors.
And these are targets, not mandates.
Aspirational targets tied to grant funding. The DOT can't force a county in rural Ohio to install RSUs. They can offer to pay for eighty percent of it and hope the county agrees to cover the rest. Some do, most don't. The NYC Connected Vehicle Pilot is the showcase — launched in 2020, expanded through 2025, over ten thousand vehicles equipped and more than three hundred intersections broadcasting. Tampa's THEA pilot tested digital work zone signs that updated speed limits dynamically based on where the construction crew was actually standing. Those are the success stories.
And outside the pilot zones?
Most intersections in America are still just intersections. They've got a stop sign, maybe a traffic light, and no radio equipment at all. The chicken-and-egg problem is brutal. Automakers won't install onboard units at scale until there are enough roadside units to make them useful. DOTs won't install roadside units until enough vehicles have onboard units to justify the expense. The current strategy to break that loop is fleet-based deployment — transit buses, delivery trucks, ride-share vehicles get equipped first. If every city bus and every UPS truck and every Uber is broadcasting and receiving, you've created a critical mass of equipped vehicles that makes the infrastructure investment worth it.
And how long does that take?
The DOT's own plan says 2036 for eighty-five percent of the National Highway System. That's the highways. City streets are a whole different timeline — there's no target for those yet. And even at eighty-five percent, that leaves fifteen percent of the system unequipped, which means physical signs are still doing the work for millions of drivers.
Which brings us to Daniel's big question. Will physical signs be completely replaced?
No. And the standards bodies are explicit about this. Digital messages supplement physical signage — they don't supplant it. The reason is liability and mixed traffic. Until every single vehicle on the road is equipped — and that means every 1998 Honda Civic, every motorcycle, every bicycle, and every pedestrian who needs to know whether to cross — the physical sign has to be there.
And "every vehicle" is what, 2045 at the earliest?
2050 if we're being optimistic. The average car on American roads is over twelve years old. A car sold today without an onboard unit will still be on the road in the late 2030s. And that's assuming every new car gets an OBU starting tomorrow, which they don't. The mandate discussions keep stalling.
So the real future isn't replacement. It's redundancy.
That's the digital twin concept. Every physical sign gets a digital twin broadcast alongside it. The physical sign remains for human eyes — and as a legal backstop. The digital message serves automated systems. Over time, as vehicle automation increases and the equipped fleet grows, physical signs may become smaller, less frequent, or consolidated. A single RSU might broadcast the equivalent of six signs that used to be on six separate poles. But they won't disappear.
Because a stop sign installed today will still be there in 2045.
The service life of a physical stop sign is twenty to thirty years. An RSU's expected service life is seven to ten. The electronics degrade faster than the aluminum. The radio modules get obsolete. The certificates expire. The physical sign just... sits there. Reflective sheeting fades eventually, but it's a known, slow degradation. You can budget for replacement on a twenty-year cycle. The RSU needs active maintenance, software updates, security patches, and someone to remember to renew the cellular data plan.
And that's where the edge cases get interesting. What happens in rural areas with no power and no connectivity?
Physical signs. That's the answer. A stop sign at a rural intersection in Montana doesn't need electricity, doesn't need a network connection, doesn't need a certificate authority, and doesn't need a technician to visit it every year. It just works. The RSU model assumes power and connectivity. For the vast majority of road miles in the United States — not the vehicle miles traveled, but the actual geographic coverage — that assumption fails.
Construction zones are another one.
Temporary traffic control is the hardest case. When a crew sets up cones and temporary signs for a two-week paving job, there's no permanent RSU. There are portable RSUs being developed — essentially a trailer with a solar panel, a battery, and a radio — but they're expensive and they require someone to configure them correctly for each deployment. The physical sign on a folding stand costs eighty dollars and takes thirty seconds to set up.
The economics of the physical sign are hard to beat for anything temporary or remote.
The physical sign is one of the most cost-effective safety devices ever invented. It has no moving parts, no power requirements, no network dependency, and a failure mode — fading over years — that is slow and visible. The RSU is better in almost every technical dimension except the one that matters for infrastructure: it's fragile.
Let me ask you about Europe, because Daniel's in Jerusalem and the European standards often influence what happens there.
The C-Roads project is the big one. Eighteen member states, running since 2017, testing harmonized V2X across borders. They've done cross-border handover tests — a car crossing from the Netherlands into Germany and the digital signage handoff working seamlessly between the two national systems. But even in Europe, where the regulatory environment is more coordinated than the U.S., physical signs remain mandatory as backup. The Vienna Convention on Road Signs and Signals still governs, and it doesn't recognize digital-only signage.
The international legal framework hasn't caught up to the technology.
Probably won't for decades. The legal system moves at the speed of treaty negotiation. The technology moves at the speed of silicon. The gap between them is where the physical sign lives.
What about Michigan's I-94 corridor? I remember reading about that one.
The Smart Corridor project. It's specifically for truck platooning — digitally linking trucks together so they can follow each other at close distances to save fuel. The digital signage broadcasts speed harmonization messages and lane closure warnings. But the physical signs are untouched. They didn't remove a single sign. The digital layer was added on top.
Which is exactly the pattern. Nobody's taking signs down. They're layering.
That's the honest answer to Daniel's question about replacement. Physical signs aren't being replaced. They're being duplicated. The digital system is a second, parallel infrastructure that does the same job for a different audience — machines instead of humans. It's not a transition from one to the other. It's a duplication, and duplications are expensive.
Who's paying for all this?
Right now, federal grants. The 2021 infrastructure bill included money for V2X deployment — not a dedicated line item, but eligible under several programs. State DOTs apply, get funded, run pilots. The long-term funding model is... unclear. The idea is that once the safety benefits are proven — fewer crashes at equipped intersections — the economic case makes itself. But proving safety benefits takes years of data collection, and in the meantime the RSUs need maintenance budgets that compete with pothole repairs.
Pothole repairs always win.
Pothole repairs are visible. Constituents call their city councilor about potholes. Nobody calls about an RSU with an expired certificate.
Hilbert: I spent six months in 2019 as a traffic signal technician for a mid-sized county in Ohio.
Go on.
Hilbert: I wasn't an engineer. I was the guy who climbed the poles to replace the bulbs. And I can tell you, the idea that we're replacing stop signs with radio waves is hilarious to anyone who's seen how actual traffic infrastructure gets funded.
What did the maintenance picture actually look like?
Hilbert: Every RSU has a power source and a network connection. Every single one. A stop sign has neither. You know what happens to the RSU when the county's IT budget gets cut and the cellular modem contract lapses? The sign goes dark. The physical stop sign still works. It's been working since 1915. The county installed twelve RSUs in 2021 as part of a federal grant. By 2024, only four were still broadcasting.
What happened to the other eight?
Hilbert: Three had certificate expiration issues — nobody was tracking the renewal cycle. Two had their cellular plans lapse because the procurement office didn't renew the contract. Two got taken out by a snowplow during the same storm in February 2023. And one just stopped working and nobody could figure out why, so it's still up there on the pole, dead, waiting for someone to climb up and look at it. The physical signs were all still there. They always are.
That's a fifty percent failure rate in three years.
Hilbert: That's a county that actually tried. Most counties didn't even apply for the grant. My brother-in-law works for a county two over from where I was. They looked at the grant requirements — the maintenance obligations, the data reporting, the five-year commitment — and passed. He said their public works director called it "a free puppy."
Free upfront, then you're buying dog food for fifteen years.
Hilbert: He's not a reliable source on most things, but he was right about that. The grant pays for the hardware and installation. It doesn't pay for the IT staff you need to manage a public key infrastructure. It doesn't pay for the cellular data plans. It doesn't pay for the bucket truck to go out and swap a unit when it fails. And it definitely doesn't pay for the liability lawyer when someone runs a stop sign and their attorney asks whether the digital sign was broadcasting that day.
Were they?
Hilbert: The four that were still working? Most of the time. There were gaps. The cellular modems would drop and reconnect. The certificates had a seven-day lifetime and sometimes the renewal server was slow. The spec says the RSU is supposed to keep broadcasting with an expired certificate for a grace period, but the vehicle's supposed to reject expired certificates. So during the gap, you've got an RSU that thinks it's working and a car that's ignoring it.
Which is worse than no RSU at all. At least with no RSU, the car knows it has to rely on cameras.
Hilbert: That's the part the engineers don't like to talk about. The system fails in ways that are invisible to the driver. You're approaching an intersection, your car is getting no digital signage, and you have no idea whether that's because there's no sign there or because the RSU's certificate expired three hours ago. The physical sign doesn't have that problem. It's either there or it's not.
The failure pattern is silent.
Hilbert: The failure pattern is a blank dashboard where there should have been a stop sign icon, and you don't notice because you're looking at the road — where the physical stop sign is, hopefully.
What you're describing is a system that requires a maintenance culture that doesn't exist yet.
Hilbert: It requires treating traffic signs like IT assets. Most counties have one IT person and they're busy keeping the payroll system running. They're not thinking about certificate renewal for intersection seventeen. And the federal government can't do it for them — there are something like three hundred thousand signalized intersections in the U.S. and millions of stop signs. The DOT can't manage certificate lifecycles for all of them.
The institutional reality is that the maintenance burden shifts from the physical sign — which is almost zero — to a networked device that needs continuous attention. And the institutions that own the signs aren't set up for that.
Hilbert: They're set up to send a guy in a truck to replace a bulb every two years. That's the budget. That's the staffing. That's the culture. You can drop all the RSUs you want from a federal grant, but if you don't change how the county funds and staffs traffic operations, you're just installing electronic litter.
The phrase "electronic litter" is going to stick with me.
It reframes the whole deployment question. The technical capability is real — the message types work, the security architecture is sound, the pilots have proven the concept. But the gap between a pilot and sustained operations is where everything falls apart.
That brings us back to Daniel's question about whether physical signs will be completely replaced. The honest answer is: not in our lifetimes. But the more interesting question is at what point a generation of drivers stops looking at physical signs because the car tells them everything they need to know. The real legacy replacement may not be the signs themselves — it's the human habit of reading them.
When the car becomes the primary interface, the physical sign becomes a backup. And backups, by definition, are ignored until they're needed. The danger is the moment when the backup is needed and the driver has forgotten how to use it.
A driver who's spent ten years letting the car handle stop signs, and then the car's RSU receiver fails, and suddenly they're approaching an intersection with nothing on the dashboard and a physical sign they haven't consciously registered in years.
That's the transition risk nobody's really solved. The standards assume the physical sign is the safety net, but a safety net only works if you're looking at it.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop for keeping this show running — and for the phrase "electronic litter," which I'm going to use in conversation and pretend I came up with myself.
If you want to hear more about the messy reality of infrastructure — the stuff under your tires that you never think about — we did an episode on the chemistry of road paint that will make you never look at a lane marking the same way again. That's episode 4802, available now at my weird prompts dot com.
We'll be back soon.
