Horizon - car technology and crashing

[Mad at urage]

The radar reflectors I'm familiar with are just dumb sheetmetal "corners" that work on the retroreflector principle. For car radar, they'd be tiny, probably the same size as optical bike reflectors.

The supertanker comparison is apt: a 1-foot wide radar reflector on a supertanker is much more visible to a radar, than the supertanker is to your own eyes. Especially at night, or in the rain/fog.

If you are looking in the right place, if the angle of reflection is right. Yachts have problems mounting radar reflectors to maximise their potential, because the yacht isn't as stable as a tanker (i.e. it leans, often 45 degrees or so). Even with 1ft reflectors mounted at the top of the mast, sailing yachts are often lost in the clutter of reflected waves; tankers tend to use much bigger reflectors than 1ft wide and often are not passive (dumb sheetmetal corners), but these require power. For those without power to devote, the market gets complicated (and has come on leaps and bounds since I was last learning about them) - see http://www.sailgb.co...dar_reflectors/.

Bikes amongst cars share characteristics with yachts amongst tankers: They are small (harder to strap things to), have limited capacity to carry power and are relatively unstable (tend to lean about a bit!). A simple sheetmetal corner would not provide a reliable image to the car's radar (similar point in the link where "Radar systems typically require a minimum of 3 consecutive 'hits' or blips on a ship's radar before it can be acquired as a target. This puts a premium not only on the strength of the return, but also on consistent coverage") because of its constantly changing orientation (exactly the same problem that yachts have).

Downfader: My contention is that if cars were given radar to potentially spot cycles, then the onus would be put on cycles to have a good radar image: That is more difficult to achieve than most realise (much money has been spent on it by the leisure yachting industry and ... look at the weight of those things!) also IMO any potential increase in "you should have had this on your bike" is A Bad Thing.

 

[XmisterIS]

I look forward to the day when we have electric cars that drive themselves. I don't really like driving, it's boring. I'd much rather get in a smooth-running, silent electric car that hooks itself up to the matrix when you switch it on and takes you wherever you tell it to go.

 

[Mad at urage]

I look forward to the day when we have electric cars that drive themselves. I don't really like driving, it's boring. I'd much rather get in a smooth-running, silent electric car that hooks itself up to the matrix when you switch it on and takes you wherever you tell it to go.

There's a way to go yet. I tried the self-park thing in the Prius: Takes ages to get to the position it likes before telling it to park itself (is quicker to just park), then it scraped the wife's car with the bumber! Oh well, that's why I didn't try it around anyone else .

 

[Arch]

Can't download the programme just now, but I presume it's all about in-vehicle safety i.e. protecting the driver and passengers after the crash...

Take away seatbelts, take away airbags, bring on the steering column spike, I say (as suggested by Mayer Hillman).

Seat belts should be compulsory for passengers, but illegal for drivers....

Plus the spike, of course.

I also felt the programme was a bit too centred on saving people after they've mucked up the driving bit - but then it's a technology programme, and 'just be more bleeding careful' isn't going to fill an hour...

 

[snailracer]

If you are looking in the right place, if the angle of reflection is right. Yachts have problems mounting radar reflectors to maximise their potential, because the yacht isn't as stable as a tanker (i.e. it leans, often 45 degrees or so). Even with 1ft reflectors mounted at the top of the mast, sailing yachts are often lost in the clutter of reflected waves; tankers tend to use much bigger reflectors than 1ft wide and often are not passive (dumb sheetmetal corners), but these require power. For those without power to devote, the market gets complicated (and has come on leaps and bounds since I was last learning about them) - see http://www.sailgb.co...dar_reflectors/.

Bikes amongst cars share characteristics with yachts amongst tankers: They are small (harder to strap things to), have limited capacity to carry power and are relatively unstable (tend to lean about a bit!). A simple sheetmetal corner would not provide a reliable image to the car's radar (similar point in the link where "Radar systems typically require a minimum of 3 consecutive 'hits' or blips on a ship's radar before it can be acquired as a target. This puts a premium not only on the strength of the return, but also on consistent coverage") because of its constantly changing orientation (exactly the same problem that yachts have).

Downfader: My contention is that if cars were given radar to potentially spot cycles, then the onus would be put on cycles to have a good radar image: That is more difficult to achieve than most realise (much money has been spent on it by the leisure yachting industry and ... look at the weight of those things!) also IMO any potential increase in "you should have had this on your bike" is A Bad Thing.

Car radar operates on a much shorter wavelength than marine navigation radar and the reflectors will therefore be smaller (plus you don't need to be spotted from 20 miles away in heavy seas and rain).

Marine radar is extremely primitive, using pulse magnetron valve technology essentially unchanged since WW2. Car radar systems are incomparably more advanced, they trace their roots from continuous-wave solid-state military radar. You don't see capable modern radar tech on commercial and recreational shipping because that market is not large enough to give sufficient return on the R & D investment required. The automotive market, in contrast, will be large enough.

Car radars have little difficulty seeing other cars, cyclists, pedestrians, animals, trees, lampposts, kerbs, etc., none of which have radar reflectors or is optimised for to give a large radar cross section. The main point I am making is that, by adding a radar reflector, a bike CAN easily appear as large as a motor vehicle - there will be no SMIDSY as far as the radar is concerned.

Basic car radar performance is a non-issue. The challenges are in the high-level control algorithms, creating a seamless user interface and legal acceptance.

BTW, whether cyclists think it's A Bad Thing counts for nothing, we are economic and political nobodies. We will simply have to deal with it the best we can if/when the technology arrives.

 

[661-Pete]

Well, if any amount of technology can create a scenario where the idiot driver who overtook me last April, and crashed head-on into another car, can be prevented from either attempting the overtake or rear-ending me ... well and good. I think we need an awful lot of new technology, way beyond what that programme offered. There were no injuries in that accident - closing speed probably about 50-60mph. At higher speeds ...

 

[Night Train]

I thought the programme was good, in terms of the new technology that is being reserched.
The car that braked itself to prevent pulling out into oncoming traffic was good but it would just mean drivers would stop looking first.

It is unfortunate that it removes even more responsibility from the driver.


Better would be a 'black box' that monitors the driver's attention and activity in relation to the action of the vehicle and other road users. It could just warn the driver before slowly and gently reducing performance to 'safe mode' until the driver stops somewhere safe and organises a retest before they can drive again. All recordings, including video of the road and the driver would be admissible evidence.

I still believe that 'accidents' are only incidents caused by human error and better training and regular retesting would reduce that somewhat. Wrapping a driver in 'cottonwool' will just increase the feeling of security and increase the potential for recklessness.

 

[Arch]

Wrapping a driver in 'cottonwool' will just increase the feeling of security and increase the potential for recklessness.

Hear hear! Stop making cars all nice and airtight and sound-insulated and comfort suspended, and climate controlled. In my old Mini, you were in no doubt whatsoever when you were doing 70mph, 30 felt scary enough!

Taking some responsibility away from drivers is no good, unless you take it ALL away, and automate the whole process. Which might be safer, until the software crashes one day, followed shortly by the car.

Reading the Radio Times precis of the programme, it said that the 'horrific' images ought to make us all think more about our driving. I don't think it was anything like that. We're immune to car crashes, there's one in every episode of Casualty. The only gore in the programme was revealed to be fake in the hospital exercise. I think we ought to be shown more REAL results of crashes, it might get into some people's heads that it's not like a computer game where you can start again....

 

[rualexander]

Regarding the talk about radar in some of the above posts, from what I understood from the Horizon programme the main collision avoidance technology seemed to be utilising GPS signals.
The computer in the car would receive GPS data including position, speed and direction from all the other vehicles in the vicinity and from this data the computer would take control of the car if a collision was imminent and avoid the collision.
No doubt bikes could also be fitted with GPS transmitters to allow the vehicles to detect and avoid them.

 

[summerdays]

I thought the programme was good, in terms of the new technology that is being reserched.
The car that braked itself to prevent pulling out into oncoming traffic was good but it would just mean drivers would stop looking first.

It is unfortunate that it removes even more responsibility from the driver.

Perhaps the system monitoring whether the brakes needed to be applied by the car, can produce results that would reflect in the insurance premiums so that if the driver is making lots of mistakes their insurance starts to increase and they get pulled for eye tests, retraining and re-sitting their test etc.

Regarding the talk about radar in some of the above posts, from what I understood from the Horizon programme the main collision avoidance technology seemed to be utilising GPS signals.
The computer in the car would receive GPS data including position, speed and direction from all the other vehicles in the vicinity and from this data the computer would take control of the car if a collision was imminent and avoid the collision.
No doubt bikes could also be fitted with GPS transmitters to allow the vehicles to detect and avoid them.

None of the solutions can ever be produced that only rely on the car seeing another car/object due to them having a receiver on it. If that was the case we all as pedestrians would also have to carry these reflectors as well, I'm less worried whether a bike would have to have them. On the other hand think of the fun you could have as a pedestrian when if you were carrying said reflector you could simply walk out in front of cars and expect them to stop automatically!!!

 

[User482]

Can't download the programme just now, but I presume it's all about in-vehicle safety i.e. protecting the driver and passengers after the crash...

Take away seatbelts, take away airbags, bring on the steering column spike, I say (as suggested by Mayer Hillman).

Should we ban cyclists from wearing helmets? (Runs and hides).

 

[661-Pete]

If a pedestrian - or any other 'uncontrolled' road user - walks out into the road within 20 feet of a car approaching at 70mph, there's no technology on Earth can save them. It's "violating the laws of physics, Cap'n". Or is it? Increase the coefficient of friction between tyre and road, so that the car can decelerate at 8g, perhaps?

No, what we need is a means of reliably identifying those drivers who are 'safe' to be on our roads, at any one time. Remember that today's safe driver may be tomorrow's lunatic: circumstances may change. But 90% of all motorists go through their entire driving career without being involved in a single KSI. If that guessed-at figure were, in fact, 100%, we'd have no roadkill. On the whole, people know how to drive safely and considerately. Some of them simply don't use that knowledge, or choose not to use it.

My record? Not for me to boast, I suppose, I've had two serious prangs in my life, both involving vehicle write-offs, but no injuries. I've been lucky. Both occasions were - I would term it - 'bad days', for me. On both days I shouldn't have been at the wheel. Maybe that is the sort of issue that should be addressed?

 

[661-Pete]

Another comment on that programme (now having watched it on iPlayer). One volunteer driver was set the task of driving along the motorway, then she was set some task designed to sap her concentration, remembering numbers sort of thing. I don't thing they followed-up that test sufficiently: they didn't show how her driving was affected by this exercise, how many more near-misses or errors of judgement she had whilst the test was under way. Instead they merely concentrated on her bio measurements - heart rate, sweating and the like.

Which is a pity, and a missed opportunity. Because it seems to me most important to know how nearly a distracted driver turns into a killer. That way lives are lost - or saved.

 

[summerdays]

If a pedestrian - or any other 'uncontrolled' road user - walks out into the road within 20 feet of a car approaching at 70mph, there's no technology on Earth can save them. It's "violating the laws of physics, Cap'n".

I wasn't really suggesting that they did walk out ... but just dreaming about the idea of turning the current pecking order on the road on its head.

 

[snailracer]

…
None of the solutions can ever be produced that only rely on the car seeing another car/object due to them having a receiver on it. If that was the case we all as pedestrians would also have to carry these reflectors as well, I'm less worried whether a bike would have to have them...

There are currently a multitude of different ideas being explored. Placing a GPS unit on every pedestrian and bike is unrealistic and most cars initially won’t have sensors that can spot them (because sensors are expensive). Adhoc mesh networking seeks to overcome this – cars that do have sensors will inform sensor-less cars that there is a pedestrian/cyclist/other motoring hazard at certain locations, other cars can then use that information in whatever way their on-board system allows, even though they did not acquire the information first-hand.