the US lays out a road safety plan that will see cars 'talk' to each other

I wish that the safety of the people outside of motor vehicles received at least as many resources (funding, research, legal) as the safety of motorists -- who are cause of these risks in the first place.

Motorists already have strong incentives to make their vehicles safer for themselves, but they have very little incentive to make things safer for people outside of their vehicle. For that reason we need better regulations and infrastructure that account for those externalities.

Oh, that's back? I thought that died with the "V2V" scheme, which was used as a justification for 5G cellular. The document [1] comes across as a solution looking for a problem. Like much of IoT.

Waymo cars do not have, or need, a vehicle to vehicle communication system. They talk to their HQ, but that's not part of the main control loops. Most of the problems Waymos have to avoid don't communicate much, if at all. Traffic cones, pedestrians, bicyclists, and people in wheelchairs have to be sensed directly. Once you have the sensing to do that, large vehicles should not be a problem. (Yes, Tesla has had trouble sensing fire trucks, semis, etc., but that's a Tesla problem)

It doesn't seem to be necessary that traffic lights and signs broadcast their status to vehicles. It's not something Waymo has asked for.

So what is this "V2X" thing supposed to be for? Surveillance, probably.

[1] https://www.its.dot.gov/research_areas/emerging_tech/pdf/Acc...

I spent two years building a module for use with intersection signal controller firmware (makes light turn red, green, yellow). The goal was to make better use of all of this extra data that we had. Sensors have improved, and all of this V2I stuff should really complete the picture. I poured my soul into that project. I really thought a lot about extensibility and ways to make it easy to accommodate new data sources. Here we had a chance to not shoot ourselves in the foot like we did last time these algorithms were visited, back in the 80's.

And then we discovered a memory leak in some vendor's V2I component which was supposed to slurp vehicle position data from DSRC and feed it to my controller. We couldn't get it to run stably in the cabinet, so the whole thing just kind of... stopped. Then they put me on some pointless project, so I quit. This was 2017.

Last year, I ran across somebody on youtube demoing my project. I got in touch, just to say hi, and he asked me for tech support--he didn't know how to configure it. I said:

> There's a readme on the SD card, plus an SDK that should get you started writing modules for various detector types.

He said:

> What SD card? I see what looks like a memory chip, but it's soldered onto the board.

They had told me I was building a research platform, and then when it finally ends up in the hands of a researcher, it turns out they've changed the hardware to make it useless for research. I was so angry. They never wanted to change anything, they just wanted some intellectual property that they could use as a bargaining chip in some way that had no impact on the intersection whatsoever.

This "V2X" business appears to be the updated branding, and it's a good idea at its core, but from what I know about the people implementing it, I'm not going to get very excited about how it'll turn out.

The article doesn't make it clear to me when the DOT talks about "V2X being deployed", what the full scope of that is – does it refer to just the physical technologies, or the lowest layers of the OSI model? Or does "V2X deployment" here mean more application-level stuff, i.e. a series of minimum requirements about what information classes of devices will broadcast to other classes of devices, with what limitations?

Without that clarification, I think the first thing readers of HN will think, justifiably, is "is all of my car's information being broadcast all the time to everything", for plenty of reasons – dragnet surveillance, disruptive attacks ranging from Flipper pranks to state actors, etc.? It's not clear whether that's true or expected of this V2X initiative.

After some quick digging, it looks like so far, it looks like only very domain-specific features have been "implemented with V2X", and will be for the forseeable future (see p7+ in [1]) – oversize vehicle complaince, pedestrian in crosswalk, blind spot warnings. How that's implemented will probably need a lot more digging.

[1] https://rosap.ntl.bts.gov/view/dot/68128

I like the modern world and its safety features. Things (cars, planes, food, etc.) are generally safer but I just do not really trust any of the people that would be writing these rules. I fear the regulatory capture aspect of it and what it might mean to me trying to get to the grocery store. I only drive a three or four times a week (most of that is short duration rural driving).

It is not that I think some one will take my car from me so much as the industry may just work to make everything not new obsolete. A new $30K car (or even $8$15K used) is a steep price for an individual to pay to meet regulations.

I really don't want connected cars. It introduces way too much remote attack surface where there was zero before.

Soon:

“Your honor, it may be true that my client’s driving speed in combination with the thick fog prevented him from reacting to obstacles, and that his car then struck and violently killed this man while he used the crosswalk. However, it was not the fog or my client’s speed that caused standard crash avoidance safety mechanisms to fail, but the crash-ee’s negligent decision to go outside without a phone with a functioning and active location beacon.”

It's kind of crazy to me that we focus on things like this, which would only work for the most modern of new cars, but don't focus on other, simple things.

In many European countries all cars need to undergo a yearly road worthiness check. So not just the emissions checks, but also the lights, tires, brakes, seatbelts, shocks, steering, and other basic things. It seems like such a system (even every 2 or 5 years) would result in major reductions in accidents and injuries and be applicable across all cars.

Self-driving cars able to communicate intent and negotiate could be extremely efficient by reducing collisions and traffic.

From a standstill, all vehicles waiting could accelerate simultaneously rather than create pressure waves due to human reaction times.

With fully-autonomous coordination, might also be possible to do away with traffic lights and other control elements to negotiate scheduling of vehicles moving across each other so they cross intersections using precisely-allocated time slots without stopping.

The number of negative comments here seems odd to me.

If you actually want practical and safe self driving cars widely deployed it seems obvious that instrumenting roads and making them a better platform for self-driving vehicles is an important part of this process.

To me this work seems like a part of the process of evolving roads from a Ad-Hoc and poorly documented system involving a lot of human guess work into a more robust and reliable platform for self-driving and human driven cars.

I see nothing about the humans outside of the metal clown cars. Meanwhile e-bikes are growing, and pedestrians and other road user die needlessly to oversized vehicles driven by anyone with a pulse due to lax licensing rules.

Fix licensing, make vehicles safe for those walking/biking not just those in other clown cars.

This won't work. None of the "safety technology" we've added in the last decade has worked. Not that the individual subsystems don't work effectively on their own, but that the entire vehicle is becoming a less safe place.

https://en.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in...

There's two main contributing factors, weight and distraction.

Weight is easy to tell by looking around. More people than ever are driving around in huge SUVs and pickup trucks.

But the bigger issue is distracted driving. The majority of other drivers I see are on their phones while moving. Almost everybody is checking their phones at intersections. And those that aren't have another distraction, the big tablets built into their cars, the use of which is required to change trivial settings.

Adding more tech won't solve this. Only removing and limiting its use will roll back this trend.

About time. I like to drive and wish to keep my liberty on the road as much as anyone else but I think it would be easy to send basic just-in-time telemetry to the other neighboring cars to improve cruise control, emergency braking and heavy traffic situations.

The car in front of you could easily send its exact speed, throttle/brake position. If it is following gps, it could broadcast the next turn on its route to help you predict its intentions (turn signals are often lacklusters or too complicated for some drivers)

In traffic, it could help stiffen the elastic by reducing the reaction time either by either telling the driver to get ready or accelerating for them.

The possibilities are infinite once you have a minimum of telemetry.

Makes sense to me. Even if it’s just a “hint” that could massively alter outcomes. eg braking a second earlier could be the difference between crash and no crash

I rode a scooter in Phnom Penh, Cambodia, for about six months. It was fascinating because the road traffic controls are largely ignored, in favour of communicating via body language with the vast numbers of other scooters around you, pretty much treating the odd car as a stationary obstacle. There's also a strong feeling of "we're all in this together and we co-operate to keep things moving and get where we need to go".

Crossing the road there is a matter of slowly and clearly walking into oncoming traffic, trusting that they will stop or go around you [0]. It kinda works the same on a scooter - you slowly and clearly do the utterly illegal thing that you want to do, and everyone will stop for you or go around you. Riding the wrong way up one-way streets, or on the wrong side of the road, is common and perfectly safe as long as you take it slow and keep your eyes open.

I think this concept of clear communication works better than the conventional system of road traffic controls. IIRC there have been tests done in the UK where removing road traffic controls led to people slowing down and communicating their intentions more, with greater flowthrough and reduced collisions.

Automating this will be interesting. Yes it'll probably work, and accidents will be reduced. But it moves us further into making the act of driving illegal. If we are all safer when the robots drive, then it's reasonable for humans to be disallowed from driving. But is the world a better place if we don't drive our own cars any more?

[0] I have had to explain this to very confused German tourists a few times. "The light is green for me to walk, why are they not stopping?" cries a worldview being shattered.

The drive down the path of authoritarian dystopia continues... as everyone is so focused on "safety" that they lose sight of what it means to live.

You know what we don't need in our cars? More alerts. Unless my car is going to have a Terminator-style HUD to put a red box around something that I need to care about, there's only so many dings and beeps I can understand while I'm trying to drive. If I have to read an alert that a pedestrian is in a crosswalk up ahead, it's immediately getting turned off.

My car already tries to tell me when it thinks I'm about to hit someone or someone is about to hit me. By the time I've looked at the screen to know, I'd have already hit them.

How many of these warnings will eventually cause alarm overload?

That's not to say I'm opposed to safety tech. I think it's a great idea. But there's no way to safely and effectively convey all that information to me while I'm already driving. If this gets rolled out as planned, I can't see how cars won't be constantly dinging and beeping and cause notification blindness, negating the whole effort.

The other problems with capturing peoples freedom of movement, is the liability it creates in insurance and or legal accountability.

It will lead to countless edge-cases that usurp normal judgement by rational drivers. Example: "The school bus stopped on the railway crossing, because some drunk in a Tesla passed out in the turning lane."

What a silly policy from naive nerd hubris. =3

If you want to know where this is heading, keep an eye on the fortunes of Cohda Wireless.

https://www.cohdawireless.com/

Cohda was started in 2004 by three very smart Information Theorists (Grant, Alexander, Rasmussen), who basically solved the problem of how to get a WiFi/OFDM signal to/from a vehicle moving at speed through a built up city. The company name is Ad Hoc, as in Ad Hoc network, spelled backwards. Cohda's original version was effectively a preprocessor between the WiFi and the antenna, that tamed the multipath channel. Back then they had some impressive coverage maps of the Adelaide CBD, whereby plain WiFi struggled to get coverage in a 100m radius, whilst the Codah system had seamless coverage out to about 1km. They were (and still are?) the leaders in this space.

My morning $0.02:

I have thought for many years that we need to make driving a part of both middle and highschool. Not merely the principles of motor vehicle operation, but the humanities aspect too.

For example, psychology, basic physics and sociology would be integral to the curriculum. It is important to view transportation as closely as possible for what it is. As conscientious driver, I do my best to be courteous and safe, for both selfish and altruistic reasons. I try to apply my understanding of traffic dynamics every time frustration is detected. It is impossible for me to drive without observing stupidity, inefficiencies and systemic flaws. Realizing that I am part of it and not an exception, I try to view others (drivers, bystanders, pedestrians, cyclists etc) with equal or greater importance to myself. I do not tailgate, unless it is a collective circumstance, eg slow high-density traffic. I heed speed limits, general laws, and remain cognizant of signs. I expect unexpected behavior and try to not react beyond necessary correction.

And I piss off a lot of drivers. Traveling the speed limit in the right lane in low density traffic, I will be tailgated or worse. Yet, while mostly driving well within legal parameters, I make good time and often end up ahead of erratic impatient drivers.

I believe that most collisions can be avoided through rational driving practices. But many are never exposed to the concept. A mere pulse is sufficient to receive a driver's license.

Traffic enforcement also seems to be more revenue than safety driven and lacks consistency, eg ephemeral speed traps.

An essay or book could be easily written on this subject. As such an integral, ubiquitous part of society, it is amazing that such minimal attention is placed upon it. The fact that so many lives are at stake seems enough to make a religion of it. We really should do much more, without sloughing responsibility onto technology and the lottery of enforcement. For me it is one of the most outrageously glaring contradictions of expressed values there is, with carnage universally and quietly accepted as collateral damage.

>The timeline for the DoT's plan extends to 2036, by which time it hopes to have fully deployed V2X across the National Highway System, for the top 75 metro areas to have the tech enabled at 85 percent of signalized intersections and to have 20 vehicle models that are V2X capable. In the shorter term, the agency aims to have V2X tech installed across 20 percent of the National Highway System and 25 percent of signalized intersections in major metro areas by 2028.

So the same government department that took around 10 just years to get adaptative headlights approved thinks this will happen in 11 years. Yeah, not going to happen this century.

I'm far more worried about pedestrians with these larger cars than most vehicle/vehicle interactions. I'd much rather move away from cars than lean into them.

If we want a system that produces safe and effective motor transportation with increasing automation, we should probably look at air travel, which is both remarkably safe and uses a large degree of automation.

How would the air travel regulators evaluate this proposal? How does this compare to similar methods in air travel?

One of my predictions of the future is that once it is shown that self driving isn't five nines reliable, countries will start installing electronic signals on roads to make self-driving reliable. And self driving will only be legal on those roads.

If it doesn't involve the cellular network authentication won't work. The killer app will be something I can put in my house that makes anyone driving by thinks a demolition derby is going on so they'll slow down.

If it does involve the cellular network it will face opposition because people don't want to buy another plan and coverage will never really be satisfactory, even if the cell phone industry is able to force every driver to buy a plan for their car.

I wish there was some beacon or something I could wear while walking that would let cars see me and not hit me or at least slow down. There are just way too many drivers with their heads down looking at their phones. I see a ton of them just blowing through stop signs at full speed every day. It would be nice if their cars would be on the look out since the drivers seem unable to.

There's little to no info on what v2x actually is and does. I suppose it's in research, but when the linked pdfs say it's proven to work... excuse me while I press x to doubt.

The real problem is the disregard for pedestrian and bike safety and I don't see anything in "v2x" that helps the situation. In fact, I just see things getting more expensive and a big honking security nightmare.

More info, from the source:

https://ops.fhwa.dot.gov/program_areas/ops-cavet.htm

There is a link to an ongoing test in the downtown area of Tampa, FL. They’ve installed lidar near crosswalks; when a pedestrian is in a crosswalk it broadcasts a “pedestrian in crosswalk” signal that nearby compatible cars hear (they’ve installed receivers in 1000 cars).

If this reduces vehicle-vehicle collisions, that will be an improvement. However, much more needs to be done to prevent pedestrian-vehicle and cyclist-vehicle collisions, and installing transceivers in pedestrians is not a realistic option.

The main issue I see is one of privacy and government control. Kill switches, speed governors, cars communicating with other cars … soon we’re on the doorstep of the same ability the CCP has to restrict transit.

Driver monitoring is the next big win, but people will not like it. More likely to be rolled out somewhere pragmatic / illiberal like China.

Just what my car needs, a government mandated attack vector. As if the crap that car companies are already forcing on people isnt enough.

YC summer ‘28 entry: V2X-connected kids shoes, that constantly broadcast the position and heading of your child to nearby traffic.

This will be gold for advertisers, insurers, and law enforcement agencies.

And poison for anyone who values or needs privacy in any way.

Well can they get cars to dim their headlights and not mess with others' collision detection?

Yeah, there are lots of interesting applications of getting other vehicles or traffic signals to favor you by sending wrong data!

There are a bunch of references to "Security Credentials Management System providers" in the DOT document. So it sounds like there will be some attempt to reject self-serving incorrect data, and even a mechanism to report and exclude "misbehaving" devices. But there are lots of gray areas and value judgements in what's allowed to be sent.

The message types are interesting to read (second link below). I thought at first that the Red Light Violation Warning was a message where your own vehicle sends out "Hey, we're blowing this red light, everybody watch out!" But I think that one is intended to be sent by the traffic signal system. But still, if you aren't in control of what your vehicle sends, what's to stop it from broadcasting data you don't want it to send, like "Hey, we're exceeding the speed limit by 17 mph"? Another value judgement that seems like it would be made at some regulatory level. Even if the messages are completely anonymous as claimed, you probably (at a personal level) don't receivers to clue in traffic police to show up further along your route, or have insurance companies try to match those with private roadside license plate reader data. At a societal level I'm sure some people would favor that, and others wouldn't.

https://www.its.dot.gov/resources/scms.htm

https://www.its.dot.gov/research_areas/cybersecurity/scms/SC...

Maybe these cars could get their own lanes, or better yet, tracks...

Come and take my analog vehicle, feds, I dare you.

Honestly a neat idea, making cars not run into each other seems like an almost trivial idea to implement if they can talk to each other.

Americans will do anything but build alternative transportation options to cars, build more efficient cities to allow for biking/walking to be more viable (car independence), or replacing useless parking garages, parking lots with scalable housing and living.

Automated vehicle transportation is a bust. Now Americans think installing a backdoor into a vehicle is going to solve the problem. Smh.

All of this while a majority of the country is suffering from intense heatwaves, increased intensity of storms due to climate change.

This is dumb and will not work and we have known it will not work for decades.

Road safety and traffic are solved problems. Look to Tokyo, Amsterdam, Copenhagen. TLDR: provide viable alternatives to driving to move non-essential vehicles and trips off the roads, physically segregate modes, calm traffic and speeds in urban spaces etc etc. This is cheap and easy to do and doesn't require new technology lol.

It's actually hilarious how much resources, time and energy otherwise smart engineers are willing to spend on solving the wrong problem. Optimizing intersections algorithms is a great example. Intersections are best optimized by not having intersections. If there has to be an intersection, make it a roundabout to randomize the flow vs the US style start-stop-start-stop grid of intersections which is GUARANTEED to gridlock BY DESIGN. Let alone that the no 1 thing is to do all the things mentioned above to take non-essential vehicles off the road in the first place.

But you can still drive a Cyber Truck.

As an American there are few times when I think the government did something awesome.

But I’m amazed they are thinking of this. This so awesome.

Plus the FAA will need to do this as we get more electric personal aircraft

Like other networks, dumb pipes (roads) and smart endpoints (self-driving cars) will serve us best in transportation. Vehicle to vehicle communication makes almost no difference to the remaining hard problems Waymo is working on. E.g. Vehicle to vehicle doesn't help a Waymo car identify and properly handle downed power lines during a snowstorm.

How about Google and Apple teaming up, taking all the data they receive from Google Maps/Apple Maps telemetry, including the destination waypoints, using it to calculate globally optimal routing for every car on the road, and then making the cars execute it? Like, sure, this may sound like a central planning caricature but we do actually have enough computing power to pull it off in this case! It will be glorious! And pedestrians can be easily taken into account since they all carry small GPS/radio-trackers on them anyhow.