LA - Louisiana: Radar Readings Disagree with Laser Speed Cam

Louisiana: Radar Readings Disagree with Laser Speed Camera
Lafayette, Louisiana police sergeant confirms that police radar readings disagreed with the accuracy of laser-based speed cameras.

Redflex speed cameraEither the radar guns or the laser-based speed cameras used in Lafayette, Louisiana are inaccurate according to police tests performed in last month. Lafayette Police Sergeant Richard Rees told the Daily Advertiser newspaper that the variance in readings in a December 7 test reached as high as 8 MPH, with the speed cameras in nearly every case showing the higher reading. The city performed a second test on December 26 in light of the discrepancies.

City officials attribute the speed differential to a physical effect known as cosine error. This refers to what happens when a radar gun is not used directly in a vehicle's path. The speed estimate will decrease as the angle between the radar unit and the road increases. The fixed speed cameras use lidar units placed with a more direct line of sight to the vehicle.

Despite the cosine effect, the speed camera returned essentially identical speeds at the intersection of Johnston and South College Road, but not at two other intersections. The city could not explain how this could happen and denied there was ever an 8 MPH difference in speed recordings.

"I never agreed that all of the Redflex speeds were accurate and in talking to the officer that operated the radar he consents to only Johnston and South College." Rees wrote in an internal email obtained by the Advertiser.

Source: Tramel: Speed-tracking cameras are accurate (Daily Advertiser (LA), 1/13/2008)

I thought laser can suffer from Cosine effects too????

The reason of "co-sine error" for the lower radar readings isn't the whole answer.
Most radar manufacturers review cosine and have graphs showing their % affect on speeds.
IF they keep 10 feet off the road for every 100 feet the vehicle (in a straight line) is away the readings will be ~ -1mph. (1)

The real issue is with how the LIDAR gun works.

LIDAR devices ONLY measure distances.
Those distances calculated out over time can determine a speed.

What makes LIDAR a problem is how its technology is implemented.

Looking at radar, the radio frequency reflected off (any surface) of the moving vehicle changes its doppler reflection back to the radar antenna.

The issue with LIDAR is where the reflection from the vehicle comes from is the problem. With radar it doesn't matter. However with the narrower beam of LIDAR, surface areas of different angles, depths and distances to the LIDAR receiver will affect the calculated readings because of influences like panning and sweep error.

Operator manuals for LIDAR guns mention to aim at a specific spot and follow (pan) with that exact spot while determining a speed reading. (2)
However, with a fixed mount LIDAR receiver, following the same point of a vehicle isn't possible.
The vehicle "walks" through the beam and depending on how reflective the light pulses are reflected / refracted back to the LIDAR receiver, the speed can be determined from any surface area reflective enough to return 40 or more LIDAR pulses in a row to the receiver (Rx lens)

These pulses are applied to a math equation of least squares (as error trapping software) and a speed is CALCULATED from the "good" returns: Calculated from the distance returns over time.

While sweep influence is not much from the front plate to the hood / wiper area of a typical sedan (~ 4 feet) is mathematically works out to be about 2.7 MPH.

Since speed is a linear graph line between to axis (x,y) commonly thought of a speed vs time
we can calculate speed v. time in feet per second:

60 MPH is 88 feet per sec a KNOWN.
30 MPH is 44 feet per sec
15 MPH is 22 feet per sec
7.5 MPH is 11 feet per sec
3.75 MPH is 5.5 feet per sec
hence the approximate
2.7 MPH is 4 feet per sec

This incorporates the distance part of the mathematics I'm talking about.

Since the beam can reflect from several locations (depths and distances) from the Rx of the LIDAR device, the speed reading can be flawed.


ADDITIONALLY:

Compounding the problem are two more items.

A vehicle not directly in line with the beam. Either at a slight angle because of changing lanes or the curvature of the road. These influences can be combined, compounding the issue. Furthermore, they can be sporadic and not consistently reproduced in "testing".
A beam which sweeps or pans along the entire side of a vehicle can influence the distance readings, since vehicles (sedans) are generally 15 - 19 feet in length.

LIDAR's LARGEST problem IS:

The "beam" is really three beams, side by side, which any portion can be reflected back to determine a distance reading. Big problem when you break it down.

Now every LIDAR gun does this in a similar fashion:
Stalker LIDAR
Kustom LIDAR (Laser Craft in GA)
Laser Atlanta
and LTI (Laser Technologies in CO)

Most traffic systems use either a Laser Craft / Kustom Pro III similar design (like in MD)
or units from LTI.

All LIDAR operates in the same fashion and the technology is influenced in the same way.

So the real issue becomes, WHICH portion of WHICH of the three beams is being reflected / refracted back to the LIDAR gun? Those influences can't be seen by the naked eye. Which brings into account the last factor:

SET UP:

Aiming by the installer.
IS the unit calibrated to have the three beams where the crosshairs of the view finder / scope (Heads up display - HUD) is aimed at a specific point? An operator has to perform an X and Y axis sweep of a police LIDAR unit to test for this "accuracy". Something the traffic system installer would have to do, since the mounting in the system box, doesn't permit this.