Meet your objective: Data acquisition in the driving simulator

In the final article in our tool integration series, we examine how data acquisition systems and biometrics tools add objective measurements to subjective feedback in the driving simulator.

From racecar setup to human factors research, there are many ways in which Cruden’s customers marry subjective impressions with recorded data in the driving simulator. For example, universities that use Cruden simulators for research are always finding new ways to draw more robust conclusions. This includes integrating biometrics tools that enable them to compare the subjective recollections of how test subjects think or say they acted, with data on what they actually did.

At a basic level, monitoring the driver or other occupants of the simulator could mean setting up a video camera to record the driver and projection screen. This not only shows what the driver did, and when, but makes it simpler during the analysis phase to hone in on pertinent parts of the experiment

Driver monitoring can also involve measuring the simple but effective technique of galvanic skin response (GSR), which measures humidity on the skin: if you’re anxious then you immediately start sweating! Monitoring may also encompass eye tracking from experts like Smart Eye and iMotions. Recent Cruden biometrics projects have included integrating iMotions tools at the Ostfalia University of Applied Sciences in Lower Saxony, Germany, which, like the University of Michigan, also uses its simulator for engineering projects. Meanwhile another biometrics simulator project, for a government institute, will develop requirements for the HMI systems used by police officers and other public servants.

“In developing systems like these, it is important to also measure how people are working with the HMI,” explains Dennis Marcus, Cruden’s commercial manager for automotive and motorsport. “What are the occupants of the car looking at? Is there a safety alert from the car? Is the visual clear enough? How much time does it take for them to notice the message? How long does it take them to read it, understand it and respond?”

As with many of the engineering tools we’ve examined in other articles in this Tool Integration series, Cruden’s SDK simplifies the integration of biometric tools with the driving simulator. Says Marcus, “When you carry out biometric measurements, you’re typically more interested in the human than the car. But you still want to know what the car is doing. For example, you monitor the steering input, because if something strange happens on the road, the driver’s natural response might be to steer slightly, even if it’s not necessary.

“Those tiny steering movements are something researchers want to measure. The same goes for the speed of the car, the load on the engine through the throttle pedal, how hard they brake and when, etc. Researchers want to measure how the person is operating the car and the HMI – which button was pressed when? We expose all of these events through our SDK such that the biometric software can access that information and timestamp it in its own environment, ready for analysis.”

Automotive OEMs are not prolific users of biometrics tools in driving simulators but frequently collect objective data on driving experiments using data acquisition systems. Correlation between physical and virtual prototypes is achieved by using recordings from the same data acquisition system to compare the performance of the virtual vehicle on a digital version of an actual road and that of the real car on the same road. A consistent driver at the wheel (ideally the same driver) ensures that inputs are the same in both cases. From there, it’s a case of tweaking the vehicle model and tuning the simulator to increase its fidelity, such that the vehicle behaves the same in the simulator as it does in real life.

“In the automotive industry, customers often have their own data acquisition software or use MATLAB Simulink,” adds Marcus. “If a driving simulator uses MATLAB Simulink as middleware – if Cruden’s Panthera SDK connects to a MATLAB Simulink environment – and your vehicle model also runs in a MATLAB Simulink environment, then you can easily connect virtual scopes and record data in MATLAB Simulink when developing and tuning models. For OEM customers that have developed their own specific data acquisition and analysis software platforms, we just integrate to provide whatever data they want in the format they need, connecting on the Panthera side through the SDK.”

To further assist in the correlation process, Cruden often creates 3D, ground-truth models of OEM’s proving grounds to use in the driving simulator. Aside from vehicle dynamics development, a well correlated vehicle model enables the OEM to conduct accurate validation of ADAS and autonomous driving systems in the safety of the simulator, which offers excellent flexibility of road layouts and traffic situations.

Motorsport teams go through a similar correlation process to hone the fidelity of their driving simulators, this time using detailed 3D track models and data recorded during practice sessions ahead of a race weekend. By comparing the trackside recordings with simulator drives, the vehicle model can be fine-tuned to match the real car before the team starts working on setup changes to make it faster.

The Panthera SDK is again used to streamline the integration of a team’s real-world data acquisition system of choice with the driving simulator. With data acquisition software often tied to a particular ECU manufacturer, popular choices for motorsport teams who use Cruden simulators include Wintax (Magneti Marelli), WinDarab (Bosch) and Pi Toolbox (Cosworth).

For more information, please contact Dennis Marcus via d.marcus@cruden.com

‍

Links to subsequent articles will be added below as they are published.

View all articles in our Tool Integration series of articles:

Article 1: Driving simulator and third-party engineering tool integration

Article 2: Four wheels good: Vehicle model integration for dynamics and more

Article 3: Hard decisions made easier: hardware-in-the-loop testing with DIL simulation

Article 4: In search of perfect harmony: HMI testing in a driving simulator

Article 5: New Panthera Corebox is at the heart of tool integration

Article 6: Standard interfaces for non-standard simulators

Article 7: Perfect Harmony: The driving simulator as a virtual OEM

‍