LTI 20/20 thwarted again.

[saaby93]

smeggy said:

saaby93 said:

think we're getting there
Your 20cm is only 8" I'm not familar with bike wheels but expected them to be at least twice that, however setting that aside.
Provided there were enough spokes at the right angle the pulses coming back could increment the distance at a rate higher than the bike speed.

Yes, but the point is "by very little" i.e. the maximum effect being a lot less than the difference between a speed limit and its associated prosecution threshold.
Hence the effect, while possible, is not significant.

I would dispute your 'very little'. On a 50mph hour bike the top spokes are moving away at a speed greater than that - the tyre surfcae would be 100mph. With a curving spoke it ought to be possible to receive back sequential distances indicating around 70-80mph
So I should be possible and significant in a few instances.
What we need is the timed photos to ballpark it

[saaby93]

Acheron said:

I cant be arsed reading all the replies, but when i use the LTI (although, we use a handheld device, which gives no photgraph) we have to conduct checks before and after usage.

We need to check the scope is aligned and do a distance check, so we then know that the laser is firing at the correct target and at the correct distance.

Its also not mentioned in the course, but i always record that there's no fresh damage on the device, and the seals over the screws are still intact, in an attempt to prove nobody has pulled the thing to pieces and been messing.

Also need to check the display, so that all numbers are showing correctly, and that a 6 doesnt show as an 8 or whatever.

Assuming all this is done, then you're pretty bulletproof im my opinion.

Im always a bit wary of trying to target bikers unless the device is on a tripod, because i find the target too small to clearly hit when it's just being held.

We can also only use them at specified risk managed sites.

Whether any of this matters, or helps, i dont know

Of course it does
It shows that if its used in those conditions it's ok
The crosshairs in the opening photo shows its not on a flat surface but could either be slip reading off the side of the bike, the spokes or anything. That's before questioning whether the crosshairs are the actual measurement point.

Pretty bulletproof isnt the same as 100%

[Acheron]

saaby93 said:

f course it does
It shows that if its used in those conditions it's ok
The crosshairs in the opening photo shows its not on a flat surface but could either be slip reading off the side of the bike, the spokes or anything. That's before questioning whether the crosshairs are the actual measurement point.

Pretty bulletproof isnt the same as 100%

I dont think you can read too much into the phot to be honest. Its one little snapshot probably taken from a video.

I think the problem in the article seems to suggest the equipment wasnt set up correctly. If the checks arent done, nobodys getting potted, regardless how fast they are going.

I think the topic title should be changed to; "LTI 20/20 Operator thwarted again".

[saaby93]

Acheron said:

I dont think you can read too much into the phot to be honest. Its one little snapshot probably taken from a video.

and thats what we're missing
You can work out approx speed from the timed photos, which'll show whether the biker was mistaken or whether the machine got it wrong.
The operator can only say what they did and the reading achieved

[Mill Wheel]

Acheron said:

I think the topic title should be changed to; "LTI 20/20 Operator thwarted again".

I thought I had explained that!
It was because the accuracy of the LTI was thwarted by an incompetent operator - just as the Cumbria case where top golfer Colin Montgomery won his appeal because the highly trained SCP operator was waving it around like a water pistol!

[Mill Wheel]

saaby93 said:

nd thats what we're missing
You can work out approx speed from the timed photos, which'll show whether the biker was mistaken or whether the machine got it wrong.
The operator can only say what they did and the reading achieved

I presume the magistrate thought of that!

[smeggy]

saaby93 said:

I would dispute your 'very little'. On a 50mph hour bike the top spokes are moving away at a speed greater than that - the tyre surfcae would be 100mph. With a curving spoke it ought to be possible to receive back sequential distances indicating around 70-80mph
So I should be possible and significant in a few instances.
What we need is the timed photos to ballpark it

No.

Perhaps I should have also described the time period of the measurement.

It may be the case that a spoke is doing 70-80mph (or 20-30mph) at one instant, but the gun doesn’t measure speed for one instant only.
It actually takes many individual samples over time: ~40 samples over 340ms.

So yes, if the bike was travelling at 1mph, you could well get a clean reading of 2mph from a spoke (but still 1mph from the tyre). However, if the bike is travelling at more than 5mph, the gun would have targeted a different spoke for a later set of samples (for that speed measurement), or the same spoke that had slowed when passing the bottom of its cycle.

To achieve the effect you describe with a lidar gun for a bike at 50mph, the gun would have to complete its measurement cycle within 40ms.
No Lidar gun operates that quickly.

The maximum possible speed error that can be attributed to this effect is 2 (or 3) mph (from the genuine speed of the vehicle).

[saaby93]

smeggy said:

It may be the case that a spoke is doing 70-80mph (or 20-30mph) at one instant, but the gun doesn’t measure speed for one instant only.
It actually takes many individual samples over time: ~40 samples over 340ms.

So yes, if the bike was travelling at 1mph, you could well get a clean reading of 2mph from a spoke (but still 1mph from the tyre). However, if the bike is travelling at more than 5mph, the gun would have targeted a different spoke for a later set of samples (for that speed measurement), or the same spoke that had slowed when passing the bottom of its cycle.

To achieve the effect you describe with a lidar gun for a bike at 50mph, the gun would have to complete its measurement cycle within 40ms.
No Lidar gun operates that quickly.

Doesnt it takes distance measurements at regular time intervals and use that to produce speed?
75mph is approx 32metres per sec and each spoke is 4cm apart
So if it took each distance reading off a different spoke what speed would the bike be travelling
Add it to PH syllabus?

[smeggy]

saaby93 said:

oesnt it takes distance measurements at regular time intervals and use that to produce speed?
75mph is approx 32metres per sec and each spoke is 4cm apart
So if it took each distance reading off a different spoke what speed would the bike be travelling

Let's say a bike was travelling at 50mph.
Let's also say the samples from the LTI were coincident with the rotation of the wheel, and were perfectly aimed, such that each laser pulse struck a spoke (only) when the spoke is at the top of its cycle. Therefore, each and every laser pulse will strike a spoke doing 70-80mph.

A radar gun in this situation would report a speed of 70-80mph. One could be forgiven for thinking a Lidar gun would yield that same result.


The LTI lidar gun would do its thing and deduce the distance to that particular spoke, for each sample. That distance would be exactly equal to the axle of that wheel; this would be the case for all pulse samples.

Therefore, the rate of change of distance from the gun to the spokes, would inherently be equal to the rate of change of distance to the axle; the latter is obviously 50mph.
Therefore, the lidar gun would display 50mph.

saaby93 said:

Add it to PH syllabus?

[saaby93]

Using round numbers
A bike is 100metres from the LTI
the LTI sends out a pulse every tenth of a second and uses that to measure distance
When its measuring the bike it gets readings of
98m 99m 100m
so the bike is moving at 1 metre in tenth of a second = 10metres per second
If instead it hits the spoke as it approaches the top it gets readings of
97 98.5 100
( ok it's a 2 metre wheel!) = 15metres per second
Very simplistic but the error in reading is much more than a few percent
Reduce time interval and size of wheel to something more realistic.

I have no idea how many spokes the OP's bike had
We need the timed photos!

Whos got a gun and a biker?

Edited by saaby93 on Thursday 5th May 19:38

[smeggy]

saaby93 said:

Using round numbers
A bike is 100metres from the LTI
the LTI sends out a pulse every tenth of a second and uses that to measure distance
When its measuring the bike it gets readings of
98m 99m 100m
so the bike is moving at 1 metre in tenth of a second = 10metres per second
If instead it hits the spoke as it approaches the top it gets readings of
97 98.5 100
( ok it's a 2 metre wheel!) = 15metres per second
Very simplistic but the error in reading is much more than a few percent

Your example works if the LTI performed only 3 samples per speed measurement.
In reality, it does at least 40 samples. Like I said earlier, it does so as a test for "stability".

So if you continue your example and calculate for the other ~40 samples, you will find your analysis will drastically fail:

a) Following that same spoke: 100 98.5 97 96.5 96 95.5 95 93.5 92 ...

b) Catching on a new spoke every 3 samples: [100 98.5 97] [97 95.5 94] [94 92.5 91] 91 ...

So at best resulting with an average speed of ...



In reality, both scenarios would result with a voided measurement due to:
1) the subsequent measurement samples not following the pattern of the first 4 samples (the first 4 are the 'acquisition phase'),
2) the lack of a smooth rate of change of distance with time.
Hence the LTI would give an error message instead of a speed reading.

[saaby93]

smeggy said:

Your example works if the LTI performed only 3 samples per speed measurement.
In reality, it does at least 40 samples. Like I said earlier, it does so as a test for "stability".

So if you continue your example and calculate for the other ~40 samples, you will find your analysis will drastically fail:

a) Following that same spoke: 100 98.5 97 96.5 96 95.5 95 93.5 92 ...

Going past 12oclock the reflections probably wrong too
If we're using an example with 40 samples choose 40 smaller intervals between 96 and 100

[JustinP1]

Going back to the OP this case has nothing to do with the merits of the LTI2020

It has everything to go with the fact that the operator had not followed the right procedure to make it worthy of supplying evidence.

I gave the same argument at court, and then at appeal at Crown court, however, in effect the Judges rules that it did not matter if or how the device is tested, neither does it have to be.

So, these cases go either way.

I was very shocked to find that despite being supposedly qualified to use the device, in my case not only did the operator not test it in the way he should, he didn't know how to test it. Further to that not only had he not followed the ACPO Code in using it, he had not read it, and furthermore he did not know it even existed.

Yet seemingly, the same guy had produced evidence which has convicted hundreds or even thousands of people and this was the first time he was brought into question.

The clerk of the court was so shocked by what he heard that in deliberation, he made a point of saying to the police officer that he really should be reading up on this stuff or he would be back here again... However, there was no need because for me they effectively supported his actions to go back and do it again.

It is my opinion that if properly questioned, maybe 50% or more occasions the device is not set up to the evidential standards set.

[smeggy]

saaby93 said:

Going past 12oclock the reflections probably wrong too

... reflecting from the other side of the spoke

saaby93 said:

If we're using an example with 40 samples choose 40 smaller intervals between 96 and 100

Scaling the parameters appropriately to get your solution to work in a real situation, you will end up with what I said yesterday (Wednesday) at 12:46:

me previously said:

Even if not, the limited radius of the spokes (let's say 30cm) would yield a maximum error of 0.3*2 /0.39 * 2.237 = +/-3mph.

Thus again yielding my same conclusion that I said today at 13:57:

me previously said:

... the point is "by very little" i.e. the maximum effect being a lot less than the difference between a speed limit and its associated prosecution threshold.
Hence the effect, while possible, is not significant.

[saaby93]

smeggy you are trying to find a general solution
and for that you have a general error margin

I'm looking at a specific case somewhere around 50mph where the advancing spoke(s) is travelling about 70-80mph, could the data back to the LTI look as though it's real data for a 75mph object.

It's like the days when watching westerns on the TV and the spokes on the wagon would look as though they were rotating backwards due to the sampling interval.

As long as the data back to the LTI looks sequential it will accept it.
It doesn't know that it's not off a homogenous surface.
On the other hand slightly outside that speed range it becomes non sequential so it discards it.

[smeggy]

saaby93 said:

smeggy you are trying to find a general solution

Nope!
I have given you the relevant specific answer to the actual, real-world scenario.

saaby93 said:

I'm looking at a specific case somewhere around 50mph where the advancing spoke(s) is travelling about 70-80mph, could the data back to the LTI look as though it's real data for a 75mph object.

It is (just) possible for a radar gun.
This is absolutely not possible for a lidar gun. I have explained why.

saaby93 said:

It's like the days when watching westerns on the TV and the spokes on the wagon would look as though they were rotating backwards due to the sampling interval.

You will notice the strobe effect never left the confines of the radius of the wheel...

saaby93 said:

As long as the data back to the LTI looks sequential it will accept it.

At 50mph: for your understanding to apply, the spoke being pinged for the later samples would have to physically depart from the wheel.

I'm afraid you are going to have to follow through your own maths (today at 19:33) with the rest of the samples, for you to understand this. If you do not then you will impede your own understanding.

Perhaps someone else can better my explanations.

[streaky]

I should have thought that the scatter from a spoke might well result in a nul or severely attenuated return at some or all of the measuring instances, which would complicate the calculation more than a tad.

Streaky

[saaby93]

streaky said:

I should have thought that the scatter from a spoke might well result in a nul or severely attenuated return at some or all of the measuring instances, which would complicate the calculation more than a tad.

yes thats what youd normally expect
What we're looking at here is whether there is a small chance that very occasionally in the right conditions the sequential data could come back in the right order for the LTI to take it as valid and put that on the screen.
Pravious posters have said t can do it when 'slipping' or when the operator waves it around.
In the OP it could be slip but could it be spokes too?

[grumpy geezer]

saaby93 said:

es thats what youd normally expect
What we're looking at here is whether there is a small chance that very occasionally in the right conditions the sequential data could come back in the right order for the LTI to take it as valid and put that on the screen.

So small that they are very close to Zero and not worth consideration when a motorcycle is targeted at any distance in the operational ranges of the LTI variants..

saaby93 said:

Pravious posters have said t can do it when 'slipping' or when the operator waves it around.

Not in the situation in the photograph and see the previous answer.

saaby93 said:

In the OP it could be slip but could it be spokes too?

No, it could be neither in that situation if the video and laser are aligned and if the operator has used the laser sights for targeting there is no chance of this either on any of the vehicles in the image.
I can imagine you becoming very excited with each new post believing you have made a discovery of a problem nobody else has spotted; not yet but I'll keep watching.

[Mill Wheel]

JustinP1 said:

Going back to the OP this case has nothing to do with the merits of the LTI2020

It has everything to go with the fact that the operator had not followed the right procedure to make it worthy of supplying evidence.

I gave the same argument at court, and then at appeal at Crown court, however, in effect the Judges rules that it did not matter if or how the device is tested, neither does it have to be.

So, these cases go either way.

I was very shocked to find that despite being supposedly qualified to use the device, in my case not only did the operator not test it in the way he should, he didn't know how to test it. Further to that not only had he not followed the ACPO Code in using it, he had not read it, and furthermore he did not know it even existed.

Yet seemingly, the same guy had produced evidence which has convicted hundreds or even thousands of people and this was the first time he was brought into question.

The clerk of the court was so shocked by what he heard that in deliberation, he made a point of saying to the police officer that he really should be reading up on this stuff or he would be back here again... However, there was no need because for me they effectively supported his actions to go back and do it again.

It is my opinion that if properly questioned, maybe 50% or more occasions the device is not set up to the evidential standards set.

Going to court is a lottery - many magistrates are as clueless as your operator, and some judges too.