All You Need
In One Single
Lorem ipsum dolor sit amet, consectetuer adipiscing elit, sed diam nonummy nibh euismod tincidunt ut laoreet dolore magna aliquam erat
Search here:

Automotive & Transportation

ADAS and Advanced Controls


AddFor and VI-Grade are involved in R&D activities in ADAS and AV field based on Driving Simulators.

NVidia Drive PX-2 is used in VI-grade Driving Simulators as on real AV prototypes with the advantage of being in the laboratory.

The use of virtual environment (software and hadware in the loop) with real ADAS and Autonomous Driving systems makes development, debugging and tuning very effective, with a huge reduction in time and cost and exploiting the possibility to exactly reproduce critical situations in controlled conditions with no risks.

This approach is very effective to directly compare different systems, evaluate new technologies, address development and study the interaction with different systems (typically with steering and brake) including the effect on driver/passenger feeling.

We use Deep Learning for two main purposes

  1. Vision: Pre-Training Vision Algorithms on Virtual Environments;
  2. Control & Feeling: Develop Behavioural-Cloning Controls on Simulator Test Driver’s Feeling to Autonomous Driving.





Sideslip Angle Estimator (SSE)




The SideSlip Angle Estimator is an algorithm developed by Addfor to estimate the Drifting Angle of a Vehicle.

The SideSlip also known as “Beta” or “Drifting” angle is the angle between the longitudinal axis of the vehicle and its speed direction. This angle cannot be measured with standard sensors but is very important for the modern Traction Control Systems.

Addfor has developed an high performance estimator based on the Deep Learning technology that outperforms the commonly known Kalman filtering approach.



Performance Traction Control (PTC)





The Performance Traction Control is an algorithm developed by Addfor to maximize the vehicle performance in every driving condition giving the vehicle the maximum available acceleration in exiting turns.

For any product details or customer specific questions our highly specialised team of Data Scientists and Engineers are available to answer you questions.                        



Virtual TPMS: tyre pressure estimator


An algorythm that allows the measurement of tyre pressures through using the sensors available in the car, without the need of specific hardware. An advanced version allows the precision of 0.1 bar exploiting the availability of GPS channels. This product has been designed to be certifiable for homologation and works with no limit of speed of application; therefore it is an effective alternative to existing TPMS.


Turbo speed estimator


We have developed a method to derive the engine turbine speed without the need of a specific sensor. This value allows a better mapping of the engine with the view to fully exploit the performance potential reducing fuel consumption and pollution, or can be used as a redundant safety check to improve the engine reliability.



High Precision Speed Estimator (HPSE)



The High Precision Speed Estimator is an algorithm developed by Addfor to estimate with high precision the speed of a racing car.

Despite the vehicle speed is a commonly available signal provided by every vehicle ECU, in racing application a much higher precision is required particularly to feed the high performance backing algorithms.

Currently most teams calculate the vehicle speed with an heuristic based on the vehicle accelerations to take in account the wheels slips. Unfortunately this approach is sub-optimal in particular in heavy braking and traction conditions where a precise estimation of the vehicle’s speed is most required.

Moreover, on vehicles that use commercial electronics derived by production ECUs, the wheel speed is affected by the correction factors applied by the electronics providers. Those correction factors are adapted for a normal use of the vehicle and usually are not documented by the ECU’s manufacturer.

When the vehicle uses a racing ECU that provided directly the wheels angular speed our system has implemented a calibration routine that finds the correct nonlinear function to transform the angular speed in longitudinal speed by taking in consideration the correct rolling radius, the wheel slips and the centrifugal deformations of the tire.

Addfor developed a high precision algorithm to estimate the vehicle speed with a maximum RMSE of 1% from 10 to 40 km/h and with an RMSE below 0.5% from 40 to 310 km/h

For any product details or customer specific questions our highly specialised team of Data Scientists and Engineers are available to answer you questions.

Technical Details

  • Speed Estimator – Racing Do Not Require Additional Sensors
  • Speed Error below 1.0% from 10 to 40 km/h (*)
  • Speed Error below 0.5% from 40 to 310 km/h (*)
  • 10[ms] Lag Guarantee with proper Data Feed (*)
  • Requires Either Tire Characterisation OR GPS signal when available

(*) Tire Characterisation Requires Typically One Day per Tire Set


Batteries State Of Charge (SOC), Health, Temperature

In the Hybrid and All-Electric vehicles the accurate estimation of the Battery State of Charge and  Core Temperature is fundamental to support the correct control strategies. We have developed a strong know-how in the development of nonlinear estimators for this kind of applications.


Virtual Flow Metering

In automotive Air Mass Meters are used for for determining the air aspirated by the internal combustion engine. We provide reliable algorithms for Virtual Flow Metering based on Ensembles of NARMAX models and Deep Neural Networks.


Driver Torque Request Forecast

Forecasting the Driver Torque Request is essential to model the correct strategies for energy saving like clutch disengagement, cylinder deactivation and other active fuel management policies. We design algorithms that learn the driver behaviour from historical data and produce an accurate prediction using the real-time data feed available on the vehicle CAN-Bus.


Electric Load Forecasting

In automotive applications the electric load forecasting is primarily devoted to optimize the overall vehicle energy management: hybrid solutions require bigger accumulators but also better power strategies to integrate the existing subsystems, like the brakes, with the new functionality, like the regenerative braking.