Motorcycles’ control logics are arising in number and complexity.
Thus, testing on proving ground and/or on public roads for verification becomes danger, hard and time consuming.
As example, the well-known ABS shall be tested on high mu or low mu with outriggers, but typically, the acceptance criteria are, at the end, limited to the subjective feeling of the expert tester.
It is hard to deep dive into what the system’s objective performances are and what the system is exactly doing. This becomes even harder when thinking about more control logics working together.
When dealing with the latest control systems, such as the Cornering-ABS or the new ARAS self-barking, radar-based ACC, the issues amplify: there is a short availability of real scenarios on which to execute calibration tests and, even when available, the dangerousness in executing the maneuvers increases and the subjective final assessment falters, bringing the rider’s psychophysical capabilities to the limits.
Moreover, with multiple target design parameters and variables, the needed time for on-road calibration would explode in days and mileage, where weather conditions may also frustrate the results.
The power of doing most of the this work in laboratory, exploiting the HIL approach, is well known, but still some limitations occur, in particular when attempting to make multiple control logics working together, like in real conditions, and when attempting to have all the physical hardware involved in the loop as well.
The idea to connect the whole motorbike enables this investigation capability and allows to reduce the riding risks and to properly verify the behavior of all the involved systems operating together.
The new brand-new solution here proposed, named OneBoxVIL for Motorcycle (OBV4M), allows testing in full safe, manned or unmanned through to the exploitation of the automation suite, decoupling systems complexity, executing hard to replicate scenarios in real on-road tests otherwise.
The final scope of the presentation is to stress the Cornering ABS and ARAS ACC implemented on board.
The target motorbike has been connected to the Real Time PC, fed by PWM wheels encoders and IMU signals computed in the virtual environment.
The High-Fidelity model of the vehicle has been implements, including proper scaled tires model for the reference rolling surface.
Fully parametrizable use cases have been implemented, enabling to test on-road realistic critical situations.
For Cornering ABS, sequences of different panic braking maneuvers have been provided by innovative braking actuation system and fault injection has been performed, verifying system self-detection capabilities and objectifying performances downgrade.
For what concerns Advanced Rider Assistance Systems (ARAS), self-braking Adaptive Cruise Control has been verified, in order to limit the induced deceleration, in combination to proper Human Interface to promptly advise the rider.
Finally, repeatability and reproducibility in testing execution is guaranteed, maintaining the same boundaries conditions (environmental conditions, road conditions, unevenness and tire-to-surface friction characteristics).
The separation of the effects allows to simplify and speed the analysis, confirming the effectiveness in performances improvement over target quantities variation.
