mr_bill

Plus V2P (vehicle-to-pedestrian) and V2B (vehicle-to-bicycle). Some folks also split out V2D (vehicle-to-device) and V2G (vehicle-to-grid).

I won't bother with the errors in the following sentences.

-mr. bill

AlmostTrick

Eventually BF will likely need a stickied "The Transponder Thread", in order to keep the heated debate out of all the other threads. There, we will be able to read stories about how "My transponder saved my life!" while others tell them transponders don't offer much protection, and reduce the number of people cycling. You heard it here first! Fun times.

Maelochs

Seems to me if the bike had a chip even if it were not moving, the car could recognize it as a bike, no matter the background, and be ready for-bike-specific actions. The rest of it ... I don't want a car anticipating my maneuvers and doubt it would be programmed to ... but it wouldn't have to identify or track my with Lidar if it knew, even before I was in its direct Lidar field, that I was coming.

For instance, a bike traveling between a row of parked cars and a couple rows of cars stopped at a light---the Lidar might not see me over all the cars, vans, SUVs ... but it would know I was approaching.

Sometimes the stuff we don't understand might seem ridiculous. Sometimes we might need to think a little differently.

Also ... lose your sense of humor around here, and on top of that attack the people who respect you .... not a great direction to be headed, eh?

Pretty sure no one of us knows everything (though i come pretty close. )

Ninety5rpm

Funniest post in the thread. Outstanding!

AlmostTrick

But seriously, even if AV's don't "need" transponders, there will be a market for them if they can offer cyclists any advantages, real and/or perceived... just like helmets.

I-Like-To-Bike

Modesty doesn't suit you.
You don't give your posts/predictions enough credit.

Ninety5rpm

Okay, but I'm still trying to figure out how they could actually help.

Say a salmon cyclist is coming from the right at an upcoming street, obscured by a building or something, so there is no sight line from the AV to the cyclist. Will a transponder help? Without a line of sight, how will the signal get to the AV? Bouncing around? Then how will the AV know where the cyclist is? GPS location info in the signal? Notoriously inaccurate, especially in urban environments. Not to mention that AVs will be programmed to enter intersections at speeds that wouldn't put a salmon cyclist like that at peril - the car will notice the cyclist immediately when sight line allows, and will be able to stop in time (or it was going too fast for that situation, by definition; easy to avoid).

squirtdad

to quote los lobos "You're always right, and I have never Been wrong "

JayNYC

If I can get a 10% reduction in getting killed or maimed when I go out for a ride, I'll take it. I cycle in Manhattan – perhaps where you cycle it's safer and you don't need it. But some of us do.

But again, transponders are at least 15 years out. I'm not going to get into a fight about something that's vaporware at this point.

In the near future it's about being seen on present-day cars radar-based collision warning systems (with the Audi A8 being the most bike-friendly). That's what we should be discussing, IMHO.

And then there's the question of how to be seen as a bigger obstacle to current-day Teslas when they're in assisted driving mode (or whatever they call it)… That's a current problem that's worth talking about.

AlmostTrick

Really? As one who believes in the fantasy of technology enabling AV's to handle any road situation and take over our roads in a few years, THIS is where you draw the line? There's no way a transponder could ever help a cyclist? I'm disappointed in you 95.

Imagine this: Faulty AV is barreling down on the unsuspecting cyclist. One or another sensor or chip has failed, or wires shorted out. Or, there is no mechanical failure, and this particular collision will be prevented "next time", by the programmers. The safety minded transponder carrying cyclist gets the message "ROGUE AV BEHIND" (via helmet mounted speaker) and is able to avoid being turned into road pizza with time to spare!

tyrion

The whole transponder issue will be determined by how well they work. If they increase "visibility"* in an AV traffic environment then they'll become popular. If not, they won't. It depends on the sensory capability of AVs.

*Increase visibility in an a safety improving range. If they increase visibility at 5 miles, but don't increase visibility within 1000 feet, they won't be that important for safety.

Ninety5rpm

Maybe we should be discussing that, but not in this thread. This thread is about safety related to self-driving cars, not human-driven cars with driver assist radar-based collision warning systems.

northernlights

Transponders on every pedestrian and bicycle? That's just the beginning. Perhaps road signs will require their own individual transponder as well, so the robocar can read it. Get ready for gas taxes, and other taxes to go up drastically. The cost of infrastructure (and its maintenance) required for this technology could be astronomical (ie: cost-prohibitive).

Ninety5rpm

By transponder on a bike, I was thinking something that emitted a signal for AVs to detect, not something on the bike to detect AVs.

The problem of a cyclist being hit from behind is way over blown with respect to human-driven vehicles; the risk of that happening with an AV is going to be a fraction of that. That's what I mean by a marginal benefit. Not reducing risk by 10% as someone suggested. But reducing risk on the order of .000001%. Not worth the complexity, cost and effort, at all.

Ninety5rpm

So let me see if I got this right.

In our current world, with human drivers that can only pay attention in one direction at a time, have relatively poor reaction time, are prone to distraction, drink and drive, etc., and kill 30,000 people per year in the US alone, most pedestrians don't bother using lights even at night, many bicyclists ride at night without lights, etc.

But in the AV world, with AVs which are capable of seeing in all directions simultaneously, see exceptionally well in the dark, are not prone to inattention, don't drink, have speed-of-light reaction times, etc, and are generally near-perfect at avoiding collisions, orders of magnitude better than human drivers, we're going to need to use transponders while walking, cycling and (presumably) driving human-driven cars?

Does this really make sense to anyone?

mr_bill

Yes.

(Clarifying, V2X makes sense. Word salads....)

-mr. bill

02Giant 

Will an AV know when a roadway suddenly disappears?

Ninety5rpm

You mean like due to sinkhole?

Probably has a better chance than a human driver to notice it.

I would certainly include those scenarios if I was in charge of simulation testing for AVs.

Especially for AVs working in Florida.

02Giant 

Sink holes, bridge collapses, washouts etc.

How will it know?

Ninety5rpm

Same way a human knows. It will see the road ahead is not there.

One of the big things they do in testing is take an actual drive and replay it in the simulator. They they alter the simulation. They can add a disappearing roadway to it, though I don't know for sure that they do that. I do know this: if they do and it doesn't stop the first time, they can teach it to stop, and even back up as appropriate, and update all cars with that learning.

Even today, much less five years from now, I'd much rather be in an AV than in a human-driven car if something totally unexpected like that happens. The AV will be on to it much faster.

Ninety5rpm

That's not what I asked. V2X makes sense for what?

I'm talking specifically about making collisions less likely. Does V2X make sense for making collisions less likely?

02Giant 

I'm skeptical, I would like to see a demonstration of it, real time.

Ninety5rpm

Why don't we see how you do in real-time in that situation?

AlmostTrick

The best human drivers can go decades without collisions. Even the more average ones often manage to go years.



I don't know why you'd prefer to be in an AV that still needs to learn just to reach that point. Now maybe someday they will be better than average drivers, but that day is not today.

And topping the most experienced drivers who have excellent, decades long records, may never be achieved.

Rollfast

What ever happened to VH1? did it go away like MTV?