Go digital.
Wheel force measurement with Virtual Sensors

Aim of wheel force measurement

Today, measurement of the wheel forces and torques of passenger cars and heavy duty trucks is usually done by using Wheel Force Transducers (WFT) from various manufacturers. Sometimes, strain gauges are used as alternative.

Wheel force measurements play a crucial role for:

  • reliability engineering in component design: assess vehicle component loads and fatigue

e.g. regarding suspension, car body and drive shaft

  • brake tests, to analyze for example what happens in case of full braking

Because of their complexity, wheel force measurements with WFT or strain gauges are currently limited to R&D processes.

What makes wheel force measurement so complex?

Wheel force measurement is complex and expensive, if you include the entire signal processing and data transmission, as well as maintenance tasks. The sensor hardware – WFT – is costly, laboratory-type equipment. It requires frequent calibration by specifically trained technicians, because the calibration factors might vary with ambient temperature. It has also a limited shock resistance.

Large scale test drives of a fleet of vehicles equipped with these sensitive hardware sensors on public roads are not realistic. Usually, most tests are done on a proving ground, with limited or only assumed correlation to real-life usage profiles.

  • How can a vehicle or equipment manufacturer collect reliable and extensive wheel force data during the R&D phase?
  • How to measure wheel forces precisely in series vehicles without laboratory-type hardware?
Wheel force measurement with virtual sensors | COMPREDICT

How does a virtual wheel force transducer work?

Our virtual sensors can measure the forces and torques at the wheels after proper training of the algorithms.

In the first step, the algorithms are trained for a limited time to match target values given by real hardware sensors, i.e. commercially available wheel force transducers (WFT).

  • Engine speed and torque
  • Wheel speeds
  • Longitudinal and lateral accelerations
  • Yaw rate
  • Steering angle, speed and torque

In a test vehicle driving on a test track, these virtual sensors are tuned for optimal correlation to the target values given by hardware sensors

Then, the virtual sensors for wheel forces and torques can be deployed in a larger fleet of R&D vehicles, while avoiding the shortcomings of unreliable WFT. As virtual sensors are softwarebased, they need no specific waterproof or anti-schock construction and work w/o adapter: the rim stays in place and nothing changes regarding the mount of the tire. The measurement principle of our virtual sensors does not depend on temperature, it does not suffer from excessive heat input. Virtual sensors need no adaptation for any similar vehicle, they are automatically compliant as long as the vehicle topology remains the same.

What is the output of a virtual wheel force transducer?

Like any hardware WFT, our virtual sensors measure the forces (N) as well as torques M (Nm) at the wheels. After training with appropriate data, we are able to achieve excellent correlation (e.g. about 150 N regarding Fx and Fy) between hardware wheel force transducers and our virtual sensors on front and rear wheels

Wheel force measurement in series cars

Our virtual sensors can be deployed in connected vehicles, so that real-life and highly reliable data can be gathered at scale. This way, every single series-production vehicle can be turned into a test vehicle.

When wheel forces and torques are measured in series-production vehicles, this opens up a broader application spectrum. It allows for instance:

  • road load data acquisition in real life for deisgn and dimensioning optimization
  • predictive maintenance of vehicle components, e.g. tire wear estimation
  • road condition detection

One of our customers, a high-end car manufacturer, gathers precise in-use data with virtual wheel force measurement in series cars. Based on such in-use data, additional services can be offered to the driver or fleet manager, e.g. in terms of predictive maintenance.

xEV vehicle directional stability

Especially for electric vehicles with independent 4WD, an exact estimation of longitudinal force is one of the key input values for lateral stability and path-following control functions.

Other related use case: steering wheel force measurement

We have also developed virtual sensors to measure the forces and torques in the steering system.

Benefits of virtual wheel force sensors

For our customer, this solution means

  • No worries about speed bumps / loss of data and invalid data due to hardware sensor defects / bad calibration
  • Extensive and reliable durability data from a whole fleet of vehicles in real-life usage
  • Centralized data management, when coupled with COMPREDICT’s AI-CORE and AI-VIEW
  • Shorter time to gather durability data in an effective manner
  • Reduced testing time and effort and space-saving in equipped vehicles
  • Dramatically reduced scaling costs in terms of hardware and operating expenses


  • No worries about speed bumps / loss of data and invalid data due to hardware sensor defects / bad calibration
  • Extensive and reliable durability data from a whole fleet of vehicles in real-life usage
  • Centralized data management, when coupled with COMPREDICT’s AI-CORE and
    AI-VIEW
  • Shorter time to gather durability data in an effective manner
  • Reduced testing time and effort and space-saving in equipped vehicles
  • Dramatically reduced scaling costs in terms of hardware and operating expenses


Get in touch with us to discuss about your Use case!

Contact

+49 (0)6151 3844614

contact@compredict.ai

COMPREDICT GmbH

Rheinstraße 40-42

64283 Darmstadt, Germany