Monday, February 14, 2011

Maryland SafeZones Cameras Not Independently Certified

Documents obtained by StopBigBrotherMD.Org show that the VITRONIC Poliscan speed cameras used by the state-run Maryland SafeZones program may have run afoul of a requirement under state law because they have not been independently certified for accuracy.
State law governing speed cameras in freeway “work zones” (21-810) requires the following
"(6)    (i)   A work zone speed control system shall undergo an annual calibration check performed by an independent calibration laboratory.
(ii)   The independent calibration laboratory shall issue a signed certificate of calibration after the annual calibration check that:
1.   Shall be kept on file; and
2.   Shall be admitted as evidence in any court proceeding for a violation of this section.
"
The same requirement exists for locally run speed monitoring systems(21-809).  However the annual calibration certificates for the SHA’s speed cameras were in fact issued by the manufacturer, VITRONIC, not by an independent lab. 

The Poliscan devices, and their annual calibration certificates, were provided by the state’s vendor, ACS State and Local Solutions.  The SafeZones program is run jointly by the State Highway Administration and the Maryland State Police.  The original RFP for the state’s speed camera program stated “Contractor must provide speed data collection equipment that conforms to the International Association of Chiefs of Police (IACP) for speed measurement, with a maintenance record to assure equipment is calibrated and properly functioning”.  However in fact the Poliscan is not on the IACP’s  list of conforming devices.  ACS conceded in their technical proposal that the devices were not IACP certified, but that they would seek to obtain IACP approval for them.  As of this date this has not yet occurred.



The SafeZone program’s RFP also says “The camera system’s speed detection technology must be accurate to within 1 mi/h and the speed measuring device must be independently certified once per year to meet program specifications.  A signed certificate of calibration shall be kept on file and admitted as evidence in any court proceedings involving vehicles that were captured violating the speed threshold.”  However the SHA 's response to a citizen making a Public Information Act request shows that the SHA is unable to either verify all of the vendor's claims or prove the system is independently certified.  While the SHA claimed that “During the process by which SHA evaluated the request for proposals (RFP) submitted for this contract, SHA, Maryland State Police, and Maryland Transportation Authority Police conducted a field evaluation to gain an understanding of the equipment and the relative accuracy for determining vehicle speeds.” the SHA’s response letter also stated the following regarding independent testing:
#4. Where any test EVER performed to independently verify the accuracy and reliability of the "Poliscan" device by any well known organizations or agencies within Maryland or the United States of America?
Neither SHA, nor its Vendor, is aware of any published documentation from an organization, or agency within Maryland, or the United States verifying the accuracy and reliability of the Poliscan device.

#5. As this "Poliscan" device is a very complicated device, very different and unlike a RADAR and/or LIDAR system used by the all police agencies all across the United States, has any well known organization or agency confirmed and/or tested the methods of measurements of this device?
Other than the evaluation described in response to item #4, neither SHA, nor its Vendor, is aware of any organization or agency within the United States that has confirmed or tested the methods of measurements of the Poliscan device. According to the SHA Vendor, the Poliscan device is currently in the process of being reviewed and evaluated by the international Association of chiefs of Police (IA CP) for potential inclusion on the IACP's qualified projects list.


#6. Has any agency or organization within Maryland and/or even the United States ever done or analyzed a "proof of concept" test to the methods, accuracy and/or reliability of this "Poliscan" device?
Please see the response to item #5.”
[…]
 

“#9. As this "Poliscan" speed measuring device is unlike any other, different than regular LIDAR/RADAR used by police agencies, is this device "generally accepted in the scientific community as a valid and reliable method of measuring speed?"
The SHA is not aware of whether this device is `generally accepted in the scientific community" for speed measurement. Please see our response in item #5. 


#10. One of the requirements as presented by the SHA for the "speed measuring device" was that this device be capable of measuring the speed of the vehicle to within plus/minus 1 (one) mph at speeds of 100mph. As I understand it, "ACS" acknowledges this and claims that the device that they use is in fact accurate to within plus/minus 1 mph at speeds of 100 mph. Please provide evidence to this fact and statement that this "Poliscan device" is actually capable of such claims.
At this time, our office is unable to provide the document satisfying this requirement. The delay is, in part, because of the sensitive nature of the document and the testing facility from which the data is to be obtained. We are coordinating the delivery of this information and anticipate providing you a certified summary or document, with the appropriate areas referenced, stating that the accuracy of the Poliscan system meets the contract requirements stipulated by SHA.


This matter has already raised eyebrows among some state lawmakers.  A staff member from the office of one state senator responded to a constituent in an email that “The legislative intent of Senate Bill 277 was indeed to have an independent party –ie, not the vendor working the speed cameras – do the annual calibration checks. I spoke with Michael Paylor with the State Highway Administration and he explained that originally, the Vendor believed they had the authority to perform the calibration checks itself.  SHA had to tell them that that wouldn’t fly, so now the Vendor is in the process of searching for an independent party to do the checks.”  The email went on to state that should the contractor fail to find an independent lab to verify these devices who had no financial stake that this would be grounds for the SHA to terminate its agreement with the vendor and seek damages.

The fact that SHA has yet to obtain such independent certification likely explains the SHA’s inability to produce documents in response to item #10 in the letter to the citizen cited above.

In addition to being part of the legally required burden of proof in court cases, the matter of independent testing could directly bear on the accuracy and reliability of the devices and therefore whether or not innocent drivers get tickets.  As an example, we previously reported that a German study found that some versions of the VITRONIC Poliscan software allowed for the possibility of incorrectly assigning violations to the wrong vehicle.  This is due to a variable-length delay between the time of the speed measurement and the time the citation images are taken, which can cause a vehicle moving in an adjacent lane in the same direction to be incorrectly assigned the speeding violation (the error was further discussed by the German motorist organization ADAC).  We inquired of the SHA about their software version and they stated the versions number of their software was “7.18.8 (also identified as 3.1.1)" and that this was last upgraded between April 29 to May 2, 2010, which is just before they began using the systems.  This software version was not evaluated in the German study, however we inquired with the study’s author (Dr Roland Bladt from the Chamber of Commerce and Industry Lahn-Dill) who listed all of the software versions known in Germany which did not include the SHA’s version.  He further stated (translated)
In Germany, there was never to my knowledge a software version 3.1.1. For other countries other software versions are created, if only because of the language for the settings to the meter or the various types of information for the speed: kilometers per hour per hour or Miles.   The manufacturer only eliminated the finding of this error from June 2010 with the new software version 1.5.5, because the error was not previously known. It is therefore possible that this error also occurs in other countries.
StopBigBrotherMD.org tried to inquire of the SHA under a separate MPIA request about whether they had any communication with their vendor regarding the above report.  Their response was that they could not comment because the report was in German (even though our posting on the topic contained an English translation and we provided a description of the problem) but that “no correspondence with the Vendor has been performed with regard to this issue” (despite the fact that citizens have written to the SHA regarding the matter previously).  Since the SHA and State Police evaluation occurred before this defect was discovered, and because the SHA claims ignorance of the problem, it seems unlikely that they would have even known to test the system to see whether a similar error can occur in the US version of the system.

VITRONIC Poliscan cameras are also being used by Chevy Chase Village, under that town’s contract with ACS.  StopBigBrotherMD.org has obtained annual calibration certificates from the Village which show that the mobile Poliscan cameras they began using in 2009 were, like the SHA’s cameras, certified by the manufacturer rather than an independent lab.  The Gatso radar-based cameras which the Village previously used were issued certificates by a California based company named “RHF Inc”.  ACS is also the vendor for Montgomery County, Baltimore City, Baltimore County, and Bowie but it is unknown to us at this time whether this model of camera is used in any of those locations or how they were certified.

The SafeZones website reports that as of 1/31/2011 there have been 415,787 citations issued under the program.  It is unknown how many of these citations were contested, however the calibration certificates would have been required as part of the evidence packages in those cases.  A defendant hoping to challenge a citation on this basis, assuming they even knew to ask for the certificates, would have needed to be well prepared to make this case, since district courts frequently do not attempt to divine legislative intent.