Safe Speed Forums

Hi all,

Has anyone seen this exact piffle before?



This was in an email from a camera partnership, forwarded to me.

The first clue to the complete bankruptcy of the idea is that you are infinitely more likely to kill a pedestrian at 33mph compared with 32mph.

It's slightly clever in as much as it allows for braking behaviour, which adds a significant degree of plausibility.

Using "KSI" is also a slightly clever choice - with real world figures of:
Pedestrian casualties in 2003...

Killed: 774
SI: 7,159
All injuries: 36,405

... making the KSI percentage: 21.8% which is in the range of the figures given.

The original source document needs to be debunked. Has anyone seen it, or seen these figures elsewhere?

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

So at a terminal velocity of 6mph two people in every hundred are killed.
This is like a pedestrian walking/running into a stationary car.

I wonder at what speed the fatalities occure? (between 0 and 6mph)

Perhaps they should limit walking the walking pace of pedestrians to avoid fatal accidents with parked cars.

Looks daft to me

_________________
Speed camera policy Kills

The simplest counter argument would be to redraw it as a table, with initial speed down the left column and reaction time along the top. Thus you pick a speed plus a reaction time and see the resulting impact speed (or not).

That would give a much more useful indication of the true relationship between speed and reactions, as a means of avoiding accidents.

_________________
CSCP Latin for beginners...
Ticketo ergo sum : I scam therefore I am!

I've seen this before, and marvelled at how stupid it sounded then. These exact figures were dished out at a 'speed awareness' course that was shown on Fifth Gear a while back...

_________________
Carl Prescott

...and another thing:

According to this, a 28mph impact results in 78% KSI rate. We know that K's are about 10% of SI's so that implies that pedestrians struck at 28mph would have an approximately 92% chance of survival.

That's very much at odds with what has been said before.

_________________
CSCP Latin for beginners...
Ticketo ergo sum : I scam therefore I am!

Nice, except the counter to that counter is that reaction time is a constant for a given driver in given circumstances.

It's going to be necessary to prove that the 30m was a bizarre and unjustified cherry picking exercise.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Not if you present it as an "inattention index" (copyright (C) JT ).

Thus you can demonstrate that engendering a step change in people's attention (ie moving along one column in the table) has a much greater effect on the outcome than a step change in the input speed.

The "inattention index" would be a simple time delay applied before braking starts. Lowest value would be the stock HC value, then this would increase as you went further right.

The explanatory text would give examples of what might cause what degree of inattention delay, eg talking to passengers, checking speedo etc...

_________________
CSCP Latin for beginners...
Ticketo ergo sum : I scam therefore I am!

The Ashton Mackay curve gives 28% likely to be fatal and 78% likely to be serious at 28mph. I think it's OK, in that the serious figures include fatal outcomes - it's the threshold of serious that they're trying to specify.

I've just uploaded the following hi res graph with mph scale prepared from the Ashton and Mackay curves.

http://www.safespeed.org.uk/aandmhi.gif

It's too wide to include in the forum, I think, and would break the formatting, at least for users at lower screen res.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Oh, I couldn't agree more with the good sense and reason. But we're suddenly bogged down in complex justifications for the role of inattention in crashes, and they can STILL fall back on: "Yes but, for a given set of circumstances with unvaried attention...". It's a good argument. But it's one you can lose in a sound bite world.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

...of course the other counter is simply to replot the same table but with a fixed speed and a varying level of attention instead of the other way round. That will show a much more dramatic variation of output and is every bit as simple and valid as the original in terms of what it tells us about accident causation.

_________________
CSCP Latin for beginners...
Ticketo ergo sum : I scam therefore I am!

Surely not! I mean, these figures clearly show that slightly exceeding the speed limit (presumably, being an area of high pedestrian activity, this would be a 30 limit at best) is no more unsafe than driving at or below it... No no no, this will not do, we simply can't have SCPs suggesting that in certain contrived circumstances it's safe to break the limit, that's our job damnit

Hmmm how often does a pedestrian suddenly leap out exactly 30m in front of a car and then wait to be run over?

What would the table look like if the pedestrian misjudged his mark and jumped out 35m in front. Or 60m Or 3m

Surely a bettyer way to redraw the table is to show how far ahead one must be able to monitor successfully for pedestrians at a given speed? Ah yes - the stopping distance.

You bet. See the graphs on:
http://www.safespeed.org.uk/braking.html

Especially this one:


(see text)

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Is there a table to show the number of fatalities of people that didn't step out in front of a car?

gizmo mentioned (though this might be anouther thread) something about walking into a stationary car. Well I run 7 minute miles and some T**T opened his car door which hit me in the throat. I got knocked to the floor and had bad whiplash. He said he didn't see me, hmm oh yer? I was the only one in a group of 30 wearing Hi Viz.

I've been looking into this whole thing and the factors that need to be considered are:

1) Driver reaction time / inattention.
2) Braking effort (affected by driver, vehicle, tyre condition, weather, hills etc.)
3) Distance to pedestrian at start of incident (can be expressed in time or distance - time is the more driver-centric view)
4) Free travelling speed
5) Adjustments to speed due to hazards (e.g. the pedestrian in question)
6) Pedestrian behaviour
7) Time taken for the pedestrian to arrive in an obstructing position. (They don't simply materialise in front of the car - they have to walk there unless they are James T Kirk or one of his shipmates - and walking there takes time and gives warning.)

Having looked at a few graphs, I thought this one was interesting:



A more realistic view is shown in this second graph, which has the same data. We're not usually in immediate conflict with a pedestrian and the large blank area represents "normal driving". The compression of the traces into a similar space tends to illustrate that they are not terribly far apart from one another:



I haven't yet fed the figures through the Ashton Mackay curves to get serious injury or fatality risks.

The original claims (first post in this thread) are incredibly cherry picked to get the desired result - changing any parameter alters the picture massively. Someone said "contrived" and that's exactly the right word.

btw, they used 1.0 second reaction, 0.68g braking and 98.4 feet to create their table.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

... as did your graph. Of course, in practice this time will be foreshortened in most cases.

Your bottom graph is very telling - and the message is clear - look ahead as far as is practicable without taking your eye off any close balls.

Yes, exactly, I only varied the distance. The total risk experienced by the pedestrian is analogous to the area under the curves. The largest effect on risk is to vary the reaction time, especially in those real world pedestrian impact cases where inattention plays a major role.

If we also consider that pedestrians walk at 4mph = 6 feet per second, we can see that in typical circumstances it'll take a significant part of a second for a pedestrian to actually get into your path.



Yeah. I'm happy that it's starting to show the true picture. Once I have the source figures for the claims, I'll work this up into a new Safe Speed page.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety