I come across this all the time: People have their insurance claims repudiated (rejected) or they face criminal charges, based on speed data acquired from so-called “Tracker Reports.”

Most modern vehicles in most countries where asset recovery is a topic of interest are fitted with some or other “Tracking System.” The system is either installed by design or installed afterwards. In countries like South Africa, where vehicle hijackings and theft is a major issue, most insurers might insist on the installation of a “VESA-approved tracking and recovery device” as a condition of cover.

The truth is that the VESA standard is rather vague, really. According to the website of the Motor Vehicle Security Association of South Africa (VESA), the following are the guidelines with regards to their standards:

VESA / ABS Category Descriptions:

  • Category A (Stolen Vehicle Recovery systems)
  • Category B (Fleet Management systems)
  • Category C (Stolen Vehicle Recovery & Fleet Management systems)

A consumer (Business or individual) is advised to at least confirm the following when making a tracking selection:

  • Can reliable recovery statistics be provided – Is this independently audited by a 3rd party?
  • Are the suppliers’ vehicle tracking products tested for compliance against accepted international standards?
  • Are the installation methods used safe, tested, and in line with the VESA installation standards as far as people and vehicle integrity and security are concerned?
  • Does the unit have the ability to do a health check, where the consumer is notified the moment the system no longer is operational?

If the consumer phones the tracking company, can he with the right authorization, request and obtain the position of the vehicle, or alternatively can the consumer do it himself?
With this in mind, you can easily see that the standard relates more to the installation and functionality than to anything relating to its use or application. The truth is that most people – including some experts called to testify to the technology in court – have a very limited understanding of how the technology works or – more importantly – its limitations.

Movement Report

GPS Tracking is nothing more than a series of GPS Location Data, tabulated for human interpretation. You can typically see data, like vehicle ID, Time, Driver ID, Status, and the part that most people focus on: Speed.

But how is this data collected, exactly? When the “Tracker Report” indicates a “speed,” how was that determined, how accurate is it and where did it come from? In order for us to answer these questions, we need to have a basic understanding of how GPS Technology actually works, what a “Tracking Device” is, and how they interact to where this report is available.

In simple terms, GPS Technology consists of three components: The Control Segment, the Space Segment, and the User Segment.

The Control Segment is made up of the various command centres and “Up-links” that feed updated data to an array of satellites and that control, monitor, and maintain them. Satellites are not totally autonomous, since their exact position and location, in space and relative to Earth, needs to be monitors and updated constantly. Without this, their accuracy will be affected. The current Operational Control Segment (OCS) includes a master control station, an alternate master control station, 11 command and control antennas, and 16 monitoring sites.

The Master Control Station is responsible for the high-level control tasks like providing command-and-control of the GPS Constellations, computing their precise locations, Uploading navigation messages, and monitoring broadcast and system integrity to ensure constellation health and accuracy while Monitor Stations track the GPS Satellites as they travel overhead, collect navigation signals, range/carrier measurements, and atmospheric data and feed observations to the master control station. It is the job of the Ground Antennas to send commands, navigation data uploads, and processor program loads to the satellites, to collect telemetry, and to perform S-band ranging to provide anomaly resolution and early orbit support.

While all this sounds very technical, it basically means that a lot of manpower and technology goes into ensuring that the system works properly, that satellites are accurately placed, and that the most current data is available to the constellation, for the purpose of providing their services.

The Space Segment consists of all the Satellites that are currently in orbit, operational and active – and it is growing and advancing constantly. As GPS technology has modernized, various countries have joined the fray and launched their own satellites. While the US started it all, by launching the original GPS System in 1978, Russia also launched GLONASS in 1982, China launched BeiDou in 2000 and the European Union launched Galileo in 2011. India Launched NavIC and Japan have now launched QZSS – both in 2018.

But all GPS Satellites are not created equal, either. As time went by, technology started to evolve and new iterations were designed and launched, improving accuracy and reliability – but more importantly – avoiding spoofing and hacking interference.

Satellite technology can effectively be separated into various “Blocks” which currently include:

  • Blocks IIA, Launched in 1990-1997, with the last one decommissioned in 2019.
  • Block IIR, Launched in 1997-2004.
  • Block IIR-M. Launched in 2005-2009.
  • Block IIF, Launched in 2010-2016.
  • GPS III/IIIF, First Launched in 2018.

As far as current developments go, the new US Space Force set the first GPS III Satellite as healthy and available for public use in January 2020. By November 5 of the same year, The Space Force and its partners successfully launched the fourth GPS III satellite into orbit. By December 2020 the latest GPS Technology – GPS III SV-04 – received operational acceptance. By January 2021, there were 31 GPS Satellites in the orbit and operational, under US control, but by November 6, 2020, there were 76 Satellites in orbit, in total, 31 of which are operational, 9 in reserve, 3 being tested, 30 have been retired, 2 were lost at launch and 1 launched on 5 November 2020. The constellation requires a minimum of 24 operational satellites, and the typical number active is 31.

As far as the User Segment is concerned, we are most interested in the “Public Consumer” use (vehicle tracking and recovery) than we are in the Marine, Aviation, Military, Industrial, Research and Education Segments. Public Consumers include the use of GPS Technology for Vehicle Navigation, Cell Phone Geolocation, Vehicle Tracking and Recovery, Racing, etc. But we need to talk about receivers – the devices that are used to receive the GPS Data.

The typical GPS Receiver installed in a motor vehicle is actually a “little box” that contains a few components. The core components of a typical GPS Tracking and Recovery Unit consists of little more than the following:

GPS Receiver or Module (including an antenna, to receive the GPS Signal).

  • A microprocessor (for interpreting the raw GPS data).
  • Power Supply (for connecting to power).
  • Alarm Inputs (Tamper prevention).
  • SIM Card (for sending data).

The device is typically installed (concealed) inside the vehicle, but it must not be installed in a location where excessive signal interference can be present. The signal, in turn, can be affected by many things and any signal interference will affect accuracy.

The things that could affect GPS Signal Quality, and therefore positional accuracy, could include:

  • Selective Availability – Intentional GPS Location Changes for military operations.
  • Ephemeris Error – Faults in the exact reported or recorded location and position of Satellites.
  • Clock Error – Imperfections in the tracking of time, by GPS Atomic Clocks.
  • Multipath Error – The path the signal follows from Satellite to Receiver, like reflecting off surfaces or buildings.
  • System Noise – Things like internal components, radio, or electromagnetic noise.
  • Antenna Phase Center Variations – Imperfections in the design of the receiver antenna or its immediate environment (car parts).
  • Ionospheric or Tropospheric Delays – Slowing of the GPS Signal due to various atmospheric layer conditions.
  • Geometric Effects – Errors in calculations and angular measurements.

Because vehicles are in constant motion, move between buildings and trees, past mountains and valleys and near high tension power wires, cell phone towers, and other radio sources and since cell phones are now present inside vehicles, the accuracy of the system can be impacted negatively, on an on-going basis.

GPS Accuracy can be considered in two ways: User Range Error and User Range Accuracy. The degree to which accuracy can be affected is not easily explained (User Range Error), but the amount of accuracy that could be achieved (User Range Error) is easier to define.

To this end, you will often see on a dedicated GPS device, such as on the Garmin eTrex, an “accuracy value” displayed. This value is often confused as being a measure of how accurate the device is at that particular time. This is not an accurate interpretation. What the device is showing is not how accurate it actually is (how close to the real-world coordinates it is located), but rather the accuracy that can be achieved if all conditions are optimal and if all signals are received directly. In order for this to be true, you would need to actually know the current quality of each satellite signal received – not only the number of signals received. In the above example, the device is connected and “locked onto” seven satellites, of varying signal strength. The accuracy of 18ft (about 5.5m) is what would be achieved if all data for all those 7 satellites (ephemeris, time, etc) are current and set properly, with no interference and no time-of-flight errors (Ionospheric or Tropospheric Delays of any kind).

To use an example that is easier to understand, perhaps, think about a person walking. We could say that a person walking with a gait that results in stride-lengths of exactly 2 feet (61 cm), walking a total of 100 strides would cover a distance of 200 feet (61m). If they are “perfect,” this would be true. But what if one leg is slightly shorter than the other? What if they step over pebbles on the way, or slow down and therefore shorten their gait a bit? While we know what their “optimum accuracy” will be, we cannot know what their final accuracy is, until we do a measurement. If we use the time they take to cover a distance, knowing this, we could estimate their speed. But how accurate will it be, without knowing exactly how they chose to walk, exactly how far they walked, or exactly what happened to them, on the way? We simply couldn’t, without a secondary effort at measurement. What if we make a mistake counting their paces? Or if they do? What if the starting position and end position are not accurately known? Perhaps they just say “I walked from the mall,” but not from exactly where at, or inside, the mall? GPS technology is prone to the same kinds of errors: It is not 100% accurate and it can be very inaccurate.

People – even court experts – often falsely believe that the “Tracking Device” somehow received “GPS Location Data.” This is patently untrue. The only thing that a GPS Receiver ever received (as part of a larger packet of non-location data, contained in a Rinex File) is actually time. In the simplest terms, the GPS Satellite, knowing its position from Ground Segment Updates, sends a signal to the receiver, along with the time that the signal was sent. This is all that the “Tracking Unit” essentially uses to determine a position.

If the time that a signal is sent is known and compared to the time of arrival, the GPS Tracker can calculate the distance between itself and the satellite. To get back to our human example, if you know Bob left his home at 1 PM and arrived at Peter’s house at 2 PM and that he (typically) walks at 5 Km/h, you know he covered 5 Km because that’s how long it takes to walk 5 Km. But unless you know where he walked from, you have no idea where Peter’s house is. But if Bob had 5 friends and they all walked from their houses to Peter’s house and if you know where all their houses are, you would be able to estimate – rather accurately – where Peter’s house is. But then they would all have to walk in a straight line since any turns and stops would affect your assumptions.

GPS Trackers essentially work in exactly the same way and are prone to exactly the same kinds of errors. Unless you truly know exactly where the satellites are, exactly what affected the signal on the way to the receiver and exactly the path followed, you will be faced with inaccurate distance data resulting in inaccurate location data.

But let’s talk about speed – which is what we are most concerned about, in this discussion. Speed, as “Tracker Reports” go is not “measured” in any way. There is no hard-wire, data-, or mechanical connection between your vehicle and the GPS Receiver. If your vehicle is spinning its wheels, your speedometer will certainly indicate a speed, but that speed will not be equally indicated on your tracker report. Equally, if your vehicle is on a high-speed train, the engine might be off, but the speed might be indicated on the Tracking Report as “Speeding” or “Overspeed.”

  • In order for your speed to be reported, the GPS Tracking System will essentially make a number of assumptions:
  • It will assume that all position data is 100% accurate – even if it is only connected to one satellite or has very weak satellite signals.
  • It will assume that all times are 100% accurate – even though it is not equipped with an atomic clock that is regularly updated.
  • It will assume that all movements are 100% accurate – even though you might be turning corners or going up or down some hills.
  • It will assume that your vehicle followed a straight line – even though you might be driving in circles.

It will therefore take the distance, calculated from one position to the next, and divide it by the time that expired between them – ignoring all errors and accepting all assumptions. If the system is equipped with a “base map” it will even “normalize” your vehicle position to place it on the nearest road (as if it is providing navigation input). In addition, if a signal is lost, it might even assume your continued position and time data, based on your historical movements. As an example, passing through a tunnel – where no GPS signal can arrive – will not always result in a “signal lost” message, depending on the system used.

In the above example, the distance will be determined as the distance between two points, measured at 60-second intervals. The tracking Device will assume a 2-dimensional model and calculate the speed, using the time it took for the vehicle to move a certain distance – 1000m in 60 seconds. This will yield a speed result of 60 Km/h. But, as is seen from the example where the vehicle is viewed from above, the car travelling in a straight line and the one travelling through bends both covered a straight distance of 1000 m, but the one travelling through the curves covered 1500m. The Tracking System will not know this, however, and still report a speed of 60 Km/h while, in reality, the vehicle was travelling at a speed of 90 Km/h. The “Tracker Report” will therefore display a value with an error of 50%.

This is a huge problem and the same error can develop in both directions: The Tracker Report can also report very high speeds that are inaccurate. Our own analysis of actual Tracker Reports, in real-world cases, revealed errors that are as high as 80%, or more, by over-estimation! The main reason for these errors lies not in the issue of time, but rather in the inaccuracy of location.

GPS Tracking Units are great tools for recovery purposes – when they are not “jammed” or removed. They can bring a recovery agent to within sight distance of a vehicle, over time. The errors do not result in vehicle locations remaining inaccurate or in vehicles appearing a substantial distance from where they actually are (unless selective availability is activated) but appear and fluctuate dynamically. As a result, vehicles can be located within a few hundred meters (or less) very easily and spotted easily because recovery agents typically have information about the vehicle they are looking for. For that purpose, it works well.

But when it comes to speed recording, the devices are not ideally suited and can produce results that are highly erratic and even preposterous, when measured against reality. For this purpose, we would advise against the use of “Tracker Reports” in isolation. In cases where the speeds reported by Tacking Devices are used, we would prefer to compare the values to the real-world evidence. Our experience has been that the reports – when properly scrutinized – are often highly suspect, prone to errors, and of little individual assistance.

While the proper analysis of Tracker Reports is more involved than I can cover in this article, this should provide a sound basic understanding of why Expert Witnesses should be properly cross-examined on their reliance on these reports, why an Expert Witness might be needed in cases where speed is relied upon, from Tracker Reports, and why the data needs to be very carefully considered and individually verified.

Stan Bezuidenhout is a Military Veteran, a former Specialist Police Serervist, an Internationally Experienced Crash Specialist and Court Expert, a widely published author, and an inimitable Trainer with more than 20 years’ experience in the Road and Transport Safety and Crash Investigation environment. Stan has made many appearances on a variety of Television Reality and News Programs, on Live Radio Shows, and at many conferences and events. After spending most of two years in the USA, Stan has returned to South Africa permanently, to share his vast and expansive knowledge and to help grow the Expert Witness Industry in Africa.

Learn more about Stan

Personal Website: www.stanfromibf.co.za
LinkedIn Profile: www.linkedin.com/in/stanfromibf/
Twitter Account: www.twitter.com/stanfromibf
YouTube Channel: www.youtube.com/c/ibfusa
Telegram Free Training Channel: https://t.me/stanfromibf