Radar cruise control in the fog

Isn't there a component of radar that's constantly on alert, regardless of cruise being on or not? If it senses you're about to rear-end someone it'll flash a warning, hit the brakes and tension the seat belts? Maybe just stick with that in fog.

 

Back in the old days, before traction and stability controls, standard cruise controls could not properly handle lost-of-traction events on wet / icy / snowy roads. Instead of immediately backing off to regain traction, they could drive more wheel spin. And with the driver mentally disengaged and caught be surprise, recovery efforts were very significantly delayed.

Many things have since changed, making for a more complex set of circumstances.

 

I posted about that in #24. Radar cruise control in the fog | Page 2 | PriusChat

The problem is, Toyota only specifies a potential speed reduction of up to 25mph. If you're going faster than 25 and don't see a stopped vehicle somehow, you'll probably still rear-end them, just less hard than if the system wasn't there. And I wonder how much it relies on the camera vs. the radar, since the forward collision warning system becomes disabled if the camera is blocked. It's likely that the amount of speed reduction is lower in poor visibility or slippery conditions.

 

I made a light weight radar reflector, put it in the middle of the road, and tested the radar safety system for myself (no fog):

Field Test of the Toyota Pre-collision System /PCS | PriusChat

I would say that it might help, and shouldn't hurt, but it *cannot* substitute for the human driver. You still need to drive slow enough that you can see and react if it doesn't. There is just too much that it can't handle.

 

It would be interesting to see how a 4th gen handles this. I wonder if it is looking for car-like objects to reduce false alarms? So far I haven't been able to find out to what extent the camera is used. For example here are some things the camera might do:

• Lane keeping assist (definitely)
• Looking at lane lines to see if an obstacle ahead is in your lane or not
• Looking for stopped cars around a curve out of view of radar
• Identifying cars/motorcycles/bikes/pedestrians vs false alarms that might be in front of the radar (stuff by the side of the road)

If I were to test it, I'd probably want to use upholstery foam blocks instead of cardboard. Or maybe a 4x8 sheet of foil faced foam insulation carefully balanced would be similar enough in size and shape to the back of a vehicle.

 

How'd I miss this? Didn't go to well...

 

It would be interesting to try a newer model, or a larger target. I did wonder if my target was simply too small. It is highly unlikely that the radar creates any kind of image or is optimized for any particular shape. A flat sheet of metal at right angles to the car is close to being an ideal reflector. The information that the car extracts from this sensor should be:

1.) Am I getting a return?
2.) Is the return doppler shifted, and if so how much?
2.) Is the time between emission and return changing, and if so how much and in which way?

All of these would come out as numeric values that it plugs into an algorithm. There would be nothing resembling an image, just a series of yes/no questions leading to an output. The car presumably plugs these values into its formula a hundred or more times a second, tracks the results, and only looks like it is doing anything when it activates.

The way that this all relates to this thread is the simple observation that the radar system can and does fail, and that when it does you don't get a lot of time to do anything.

 

The whole point of a scientific experiment is to reduce the size of your ignorance. It worked just fine...

 

A perfectly flat reflector, slightly misaimed, is not so good. A cube-corner reflector automatically takes care of the aim by sending a very strong reflection back parallel to the incident beam.

 

Pedestrians, don't leave home without one.

 

Absolutely! At night or dusk, I don't walk on the roadside (few sidewalks in my neighborhood) without these (optical version, not radar) unless I have a flashlight or bike blinky.

Though as automatic collision avoidance features become more common, maybe I should start carrying a radar version too.

 

Retroreflectors work the same way for light. A town where I used to live would give out reflectors (colored plastic like driveway markers, but on a string) and reflective snap bracelets to everyone to make pedestrians more visible in the winter when it's dark all the time. They were kind of fun to wear.

 

They're micro cube-corner reflectors? Interesting.

 

P

Plastic ones basically are. I think flexible reflectors use tiny glass spheres.

 

IIHS rated the new Prius superior for vehicle to vehicle front crash prevention.
2019 Toyota Prius
They test at 12 and 25 mph, which likely faster than most on a highway in fog.
It doesn't say anything about avoiding pedestrians.

 

do not know about fog, but driving from PA to CT last Saturday I was in heavy rain, with lots of water being thrown up by passing vehicles. I had DRCC on, not as a replacement for proper low-visibility driving technique but more as a belt-and-suspenders thing. I noticed that at times when there was a lot of water hitting the windshield the DRCC would suddenly lose the car in front of me. Eventually, I decided that a system that wasn't performing consistently wasn't worth using, so I turned it off.

The points being:

1) Do NOT rely on DRCC or any other automated system to replace good driving technique.
2) Know the design limitations of the automation.
3) Keep an eye on what your automated systems are actually doing, and disable them if they aren't doing what they are designed to do.

 

Apologies for the trite expression, but just seems a perfect fit here:

It sounds like there's a danger of being lulled into false sense of security.

 

I'm very much old school on this, in that any form of CC is bad news when driving in compromised traction conditions such as rain, let alone heavy rain or snow or ice.

Granted, traction control and vehicle stability control in modern cars will save many situations that would have exacerbated traction loss with old-fashioned dumb cars with dumb CC. But still, it won't save them all, as we don't yet have cars smart enough and certified for autonomous driving. So xxCC is still a rotten idea, and potentially deadly.

 

I've used DRCC many times in dense fog when I felt comfortable going at least its minimum required speed - our two Toyotas have the older DRCC version that doesn't function under about 25 mph. I don't rely on DRCC to prevent running into something but I do find it interesting that it can detect vehicles driving in front of me that I can't see.

I'm far more concerned about getting hit from behind. It's not an exaggeration when I say that every time I'm creeping along in the "slow lane" in low visibility conditions, idiots in the "passing lane" are screaming past me flying blind at high rates of speed.

Some of the European cars we've owned have come with rear fog lights. I've mounted aftermarket rear fog lights on a number of vehicles that didn't come with them. I always buy "real" rear fog light units that have a European approval mark, e.g. E1=Germany, E2=France, etc.

The rear fog lights on our two current Toyotas were purchased on eBay Germany. They are small and barely noticeable when not in use - an image showing them is attached.

I've sometimes noticed when using a rear fog light in heavy fog that the same vehicle has followed me for many miles. I don't know if the driver behind me is doing it for his convenience or what but at least he might be the one who is going to get hit by a speeding driver instead of me.

 

What color? What pattern do they project?

He figures that if there's something in the road, you'll find it for him.