Car Mileage: Why GLONASS and Odometer Diverge? Part 1. Introduction

___

_

… the task of measuring the vehicle mileage is important not only for personal use, but also for transport companies and enterprises with vehicle fleets that keep track of the performance of their vehicles. Mileage is one of the main indicators of the vehicle’s condition, determining the regulations for maintenance, replacement of consumables and spare parts, as well as cost management at the enterprise. In fact, for most of these tasks it is more correct to use engine hours, but this topic is beyond the scope of this article.

There are currently two main methods for measuring a car’s mileage: mechanical or electronic odometers and satellite navigation systems such as GPS and GLONASS. An odometer is a traditional device that records the distance traveled by a car using wheel or transmission rotation sensors. However, its accuracy depends on many factors such as tire size, tire pressure, and possible structural defects in the car. At the same time, satellite navigation systems (GNSS) offer an alternative way to measure mileage, which allows you to determine the location of the car and calculate the distance traveled based on its coordinates.

At first glance, it seems that navigation and odometer should give identical readings, but in practice, discrepancies between these methods constantly arise. These discrepancies can vary from minor to significant depending on the operating conditions of the car, the density of buildings, the terrain and the technical characteristics of the navigation equipment. For example, in mountainous terrain or dense urban areas, the GPS signal can be lost or receive multiple reflections from buildings, which leads to errors in readings. The odometer, in turn, depends on the technical condition of the car, in particular on tire wear, and can overstate or understate the distance traveled.

The purpose of this series of articles is to examine the causes of discrepancies between odometer readings and satellite data, and to propose methods for adjusting the data to improve measurement accuracy. We will also discuss how a combination of odometer and navigation data can be useful for more accurate mileage monitoring, which is key to improving fleet efficiency, cost control, and maintenance planning.

The mechanism of the odometer

The odometer mileage in a car is calculated based on data about the rotation of the wheels or rotation in the transmission, which are recorded by sensors. Methods for calculating the odometer mileage are divided into mechanical and electronic.

In mechanical odometers, the mileage is calculated by connecting the car wheels to the device through a system of gears and shafts. Each revolution of the wheel transmits a certain number of revolutions to the odometer gears. Since the wheel size is known, the distance traveled can be calculated by multiplying the number of revolutions by the wheel circumference. This method was used in older car models and is practically not used in modern systems due to limited accuracy and reliability. Nowadays, mechanical odometers can be found, for example, among cycling enthusiasts.

In modern cars, mileage is recorded using a speed sensor mounted on the transmission or drive shafts. This sensor usually uses a magnetic principle and counts the number of revolutions. The data is then transmitted to the electronic control unit (ECU), where it is multiplied by the circumference of the wheel to calculate the mileage.

Some vehicles calculate mileage based on data from the ABS (anti-lock braking system), which has speed sensors on each wheel. This data is sent to the ECU, which determines the total distance traveled. Using data from all wheels helps compensate for errors when driving on curved sections of the road or when the wheels slip, which is especially important for all-wheel drive vehicles.

An example is the 7th generation Toyota Celica, which was produced from 1999 to 2006. Even within the same model, different methods of recording speed and mileage were used for different markets. For Japanese Domestic Market (JDM) cars, ABS sensors were used on the front drive axle, while for European market cars, a speed sensor in the transmission was used.

The main reasons for the error:

1. Tire wear. With wear, the wheel radius decreases, which affects the accuracy of the odometer. Therefore, it is necessary to monitor the condition of the car tires and replace them in time.  To monitor the technical work of commercial vehicles, the Waliot fleet management platform has a maintenance module.

2. Tire pressure change. Reducing the pressure increases the contact area of ​​the wheel, which can affect the accuracy of measurements.  The Waliot platform offers a ready-made solution for monitoring tire pressure in real time with automatic notifications to the driver and dispatcher.

3. Wheel size change when replacing wheels. If you install tires and/or wheels of a different diameter, the odometer reading may be distorted because the system does not account for the size change without recalibration. Although this change may seem minor, on modern vehicles manufacturers may include calibration factors to account for tire wear or the installation of wheels of a different size.

Let’s look at an example using the same Toyota Celica. In the official owner’s manual for the car, the tires and wheels installed at the factory were 205/55R15, where 205 is the tire width in mm, 55 is the ratio of the tire aspect ratio to its width in percent, and 15 is the disk diameter in inches. Let’s calculate the sidewall height of the tire: 205 mm * 0.55 = 112.75 mm. And now the diameter of the entire wheel as the sum of the disk diameter itself (we convert it from inches to millimeters) plus two sidewall heights: 15 * 25.4 + 2 * 112.75 = 606.5 mm. And finally, we can calculate the circumference of such a wheel: 3.14159 * 606.5 = 1905 mm. Thus, for one revolution of the standard wheels on this car, the odometer will add 1.905 meters to the mileage.

Additionally, the same vehicle manual contains the acceptable tire and wheel sizes for installation. The largest of them is 225/35R18. Let’s calculate the wheel circumference using similar formulas. Sidewall height: 225 * 0.35 = 78.75 mm. Entire wheel diameter: 18 * 25.4 + 2 * 78.75 = 614.7 mm. And the distance traveled in one wheel revolution: 3.14159 * 614.7 = 1931 mm. It turns out that when installing the maximum acceptable wheel size, the odometer begins to underestimate the mileage by (1931 – 1905) / 1905 = 0.014 = 1.4% or 14 meters per 1 km of travel. Here we took into account “ideal wheels in a vacuum.” In practice, if the wheels are also flat and worn out, there will be an additional error in the odometer. And this is an error with a difference in wheel diameters of only 614.7 – 606.5 = 8.2 mm!

Thus, odometer reading is a calculation based on the number of wheel (or shaft) revolutions multiplied by the circumference, adjusted for factors that may affect the accuracy of the calculation.

How does satellite navigation work?

Satellite navigation systems such as GLONASS (Russia) and GPS (USA) operate based on signals sent from satellites orbiting the Earth. The basic principle of their operation is to determine the current location of an object using data from several satellites, which allows for the calculation of coordinates and movement in real time. 

Here’s how these systems work and their features for calculating the vehicle’s mileage.

1. Satellite network. GPS and GLONASS satellites move in stable orbits and send radio signals to Earth that contain data about the time and coordinates of the satellite.

2. The principle of triangulation. The navigation device in the car receives signals from several satellites simultaneously. By comparing the time the signal was sent with the time it was received, the device determines the distance to each satellite.

3. Calculating coordinates. Based on the data received, the device calculates the exact coordinates of its location. This process can be repeated several times per second, which allows tracking the movements and changes in direction of the car.

4. Mileage calculation. The navigation system determines the mileage by summing up the distance traveled between successive coordinates, which allows you to obtain the total length of the route.

Features of using satellite navigation to calculate mileage:

1. Dependence on environmental conditions. In open spaces, GPS and GLONASS systems provide high accuracy of coordinate determination, which allows obtaining fairly accurate data on the distance traveled. In densely built-up areas, in tunnels, under bridges or in forests, the signal may be weakened or absent altogether, which leads to data loss or errors in their interpretation. In urban conditions, the signal may be reflected from buildings, which creates a “multipath” effect – a situation when a signal from one satellite arrives at the navigation device along several paths with different delays. This leads to errors in calculating coordinates and distorts the mileage. 

2. Quality of the GNSS receiver hardware and software. Navigation devices often use smoothing algorithms to improve the quality of data when the signal is weak. This can lead to slight distortions, especially when turning or changing direction abruptly. Most consumer GPS devices update at a rate of 1 to 10 measurements per second. At high speeds or with abrupt changes in trajectory, the accuracy of the calculation may decrease, since the data is not updated instantly. The higher the quality of the GPS/GLONASS receiver, the less likely it is to make errors. However, many car navigation systems use receivers of average accuracy, which can also introduce distortions into the mileage.

Thus, satellite systems provide high accuracy, but depend on the surrounding conditions and the quality of the equipment itself. At the same time, satellite systems do not require calibration depending on the size of the wheels, their wear or tire pressure. 

Real mileage of the car

We’ve sorted out the odometer and satellite mileage. But what is the actual mileage of a car? Technically, the mileage of a car is the distance traveled by its center of mass. It is the movement of the center of mass of the car that is least subject to deviations from the actual distance traveled, while the mileage of individual wheels can differ significantly. The center of mass of the car is somewhere in the middle of its structure and describes the trajectory around which all elements of the car rotate. This mileage indicator is the most accurate, since it is measured as the distance traveled on the road and does not depend on turns and wheel slippage.

Differences between the mileage of individual wheels and the mileage of the center of mass:

1. When turning, the outside wheels travel a greater distance than the inside wheels. For example, when turning right, the right (inside) wheels will travel a shorter distance than the left (outside) wheels. The turning radius for the outside and inside wheels will be different, so depending on the intensity of the turns and the trajectory, the difference in wheel travel can accumulate.

2. Slipping and skidding. When accelerating or braking hard, the front and rear wheels may slip or skid. This causes one or more wheels to travel a greater distance. In off-road conditions, slipping can significantly increase mileage, especially for the drive wheels, which does not reflect the actual distance traveled by the vehicle.

3. Differential and independent suspension. The differential system allows the wheels to rotate at different speeds when turning, which is especially important for all-wheel drive. This helps reduce tire wear and improves the vehicle’s cross-country ability, but results in a difference in mileage between the wheels on each axle. The ABS and ESP systems use the data from the speed sensors of each wheel and independently apply braking force to each of them.

How much can the mileage differ? In real conditions, the difference can be from a few percent to several tens of percent, depending on the driving style and operating conditions. On long straight trips (for example, on the highway), the wheel mileage will be very close to the mileage of the center of mass. In city traffic, with frequent turns and stops, especially on small radii, the wheel mileage relative to the center of mass can differ by several kilometers per trip, depending on which maneuvers are performed more often.

Theoretical error estimate

The discrepancies between the odometer and the satellite navigation really depend on many factors, and their extent varies depending on the operating conditions of the car. Below is a theoretical estimate for two different cases (a semi-trailer on the highway and a passenger car in the city) based on personal experience, which largely coincides with the generally accepted opinion on this topic.

For large vehicles moving along the highway at high speed with a relatively stable trajectory, the discrepancies between the odometer and GPS/GLONASS are usually minimal. On a flat road with stable tire pressure and minimal turns, the odometer may give an error of about 0.5-1% due to tire wear and possible changes in their diameter (for example, with an increase in load). For 100 km, this will amount to a 0.5-1 km discrepancy. For satellite systems on an open road, the error is also low, usually does not exceed 1%. On a route without multiple bends and intersections, the error will be approximately 0.5-1 km per 100 km. However, temporary signal losses, for example when driving under bridges or through tunnels, can temporarily introduce additional fluctuations, but they smooth out over a long distance. As a result, the overall discrepancies in readings between the odometer and GPS/GLONASS per 100 km can amount to up to 1-2 km in total. This is less than 2%, which can be considered insignificant for the purposes of accounting for mileage on highway routes.

Urban conditions significantly increase errors in both the odometer and GPS due to frequent turns, acceleration, braking and possible wheel slippage. With frequent turns (for example, at complex intersections or junctions), the inner and outer wheels travel different distances. This leads to an additional error, which can reach 1-3% per 100 km, which is 1-3 km of discrepancy. In urban conditions, the GPS signal is often reflected from buildings, and in densely built-up areas and under trees, multipath effects occur. Navigation systems can lose signal and fill in the gaps by smoothing the data, which also adds error. In the city, the error can reach 3-5% per 100 km, that is, up to 5 km of discrepancy, especially if the car often finds itself in areas with a poor signal. As a result, for a passenger car in the city, the total discrepancy per 100 km between the odometer and GNSS can be up to 5-8 km. 

It turns out that in highway conditions, the discrepancies will be minimal and will rarely exceed 1-2%, which can be considered acceptable. In urban conditions, especially on short sections and with frequent stops, errors can reach 5-8%, which is a more significant deviation. Additionally, for the southern and southwestern parts of Russia, it is imperative to take into account the use of electronic warfare (EW) and jamming of the satellite signal, which impose a significant error on the operation of GNSS in these areas.

Methodology for calculating the actual error

In my practice, I have encountered companies that asked for evidence of the error of the standard odometer and the accuracy of the satellite navigation. Most often, these were questions from the accounting department, requiring waybills and payrolls with ideal values ​​​​as on the “dashboard” without any deviations. As evidence, you can offer a simple experiment that will show the differences between the odometer readings, GLONASS and the actual distance traveled. 

1. Experimental route

Choose a route of 10-20 km for a city trip and a similar route for a highway trip to compare the impact of different conditions. In urban conditions, the route should include several turns, stops and different speed modes (traffic jams, traffic lights and free streets). In the case of a highway, choose a flat, straight section with minimal overlaps. The test route should be as close as possible to the real operating conditions of the vehicle.

2. Data collection

Record the initial and final odometer readings and/or use the “Trip” function to calculate the current trip mileage. This will show the mileage calculated by the car’s internal system. Turn on the navigation device so that it records the driving data and sends it to the telematics server of the vehicle monitoring system (VMS). Make sure that the device is correctly installed and configured, and that it has a good satellite signal.

3. Measuring the real distance

Use a high-resolution satellite map or specialized applications (for example, Yandex.Maps with manual route marking). Draw the route on the map and measure its distance manually, using the exact geographic start and finish points. Additionally, you can use kilometer markers (poles) on the route, but it is worth considering that the accuracy of such markers is usually within the error range of 0.1 to 0.5%.

4. Calculation of error

Once the experiment is complete and all data has been collected, calculate the percentage error for each mileage measurement method relative to the measured route:

* Odometer Error = (Odometer Mileage – Actual Distance) / Actual Distance * 100%

* GLONASS error = (GLONASS distance – Real distance) / Real distance * 100%

These calculations will give a clear picture of how much the odometer and GLONASS data differ from the actual distance traveled in urban and highway conditions. Obtaining such practical data will clearly show the differences between the odometer and navigation and their dependence on driving conditions.

By conducting a number of such point measurements, we will gradually approach approximately the same error figures that were indicated earlier in this article and are considered generally accepted. But for most, such evidence is not enough, and I can agree with this! Therefore,  in the next article,  we will move on to analysis on real data from the Valiot fleet management platfor.

_

_