New Trends in Driving Simulators:
The out-of-the-loop experience

Driving simulator for research purposes are used in the area of human factors and medical research, to monitor driver behaviour, performance, and attention and in the automotive industry to design and evaluate new vehicles or new advanced driver assistance systems. It is a tool that enables a person to experience driving in a safe and controlled virtual environment.

Typically, a driving simulator consists of a physical mockup, with driving commands (steering wheel and pedals) and a visualization system. The hardware configuration of these two components ranges from affordable desktop setups to complex installations consisting of, e.g., a 360-degree projection dome surrounding a real car and built on top of a moving system, often larger than a tennis court. The latter systems require dedicated IT and logistics infrastructures, with multi-million investments and high running costs. This big technological investment is undertaken to provide higher fidelity, immersion and realism to the third component of a driving simulator: The human driver.

From vehicle simulator into Driving Simulator

The primary purpose of a driving simulator is to let its user feel like in a real car and behave consequently and appropriately to the driving context. In other words, the driver eventually turns a vehicle simulator into a driving simulator. Nowadays, simulators have a well-established role along the entire vehicle design process: From the early assessment of virtual prototypes to the validation of production-ready solutions. Indeed, modern driving simulators allow designers and engineers to quickly implement virtual models and test them in highly realistic environments with selected drivers. The agility of this process, the standardization of the assessment procedures and the power of integrated software/hardware tools are surely a key factor in the success of driving simulator technology in industrial research.

Driving Simulator as a human-centered tool

The massive digitalization of industrial processes and the development of human-centered systems engineering are bridging the gap towards the design of usable, understandable, and natural Human-Machine Interfaces which can be easily implemented and tested in driving simulators. Moreover, the development of advanced Driver Assistance Systems is also being successfully tackled using driving simulators. However, the topic of Automated Driving is spreading in every sector of transportation technology and the questions have raised: Do we still need driving simulators, if vehicles are going to drive autonomously? This seems a legitimate question if one considers that the driver will soon be out of the loop, while driving simulators have been designed to bring the driver in the loop. The truth is that driving simulators are now more than ever necessary to put drivers and passengers in conditions that closely resemble the real road environment to evoke, measure, model and understand the complex relationship between humans, automated vehicles and road infrastructure.

New driver roles and Trends

A driving simulator conceived for research and development in the field of automated driving must consider the new roles that drivers are assuming. Indeed, with increasing level of automation drivers will be able to gradually phase-out control over the driving functions and transfer it to the automated vehicle. This trend will soon enable a variety of non-driving tasks that until now were simply not part of the driving experience. Moreover, safety-critical issues arise in transfer-of-control scenarios, where the automated system and the human driver need to effectively communicate to each other their intensions and actions. Therefore, a simulator must enable the study of the interactions between drivers, passengers, vehicles and road users. In line with this, it is foreseeable that the future development of driving simulators dedicated to research in automated driving will focus more on enabling onboard connectivity, driver monitoring and interaction concepts and technologies, rather than developing further motion systems. Indeed, the variety of motion-based driving simulators available nowadays already offer realistic motion cues for most driving scenarios. The opportunity of further development is rather limited, due to the insurmountable boundaries posed by a confined workspace. On the contrary, driving simulators need to be re-conceived as living spaces where humans act out-of-the-loop, in connection with each other and using different technologies, and upgraded with multi-sensory interfaces that have yet to be conceived. This perspective will have profound implications in the development of future simulator technology.

Our Research Aspects

Following the above-mentioned considerations, at Virtual Vehicle we have installed a new semi-static driving simulator, equipped with a high-end immersive visualization system and a real car cockpit. We have customized the system for research purposes with additional hardware and software components. We have integrated an eye-tracker, physiological sensors and webcams for monitoring the driver state, active belts and active seat for somatosensory stimulation to provide quick acceleration onset cues to the driver, an accurate sound reproduction system, and several mounting points for installing displays in different positions inside the cockpit.

Additionally, the simulator features a first-of-its-kind integration of the car cockpit on a tilting platform, which enables the accurate reproduction of inertial cues and vibrations for drivability and comfort studies, with no need for large motion systems. Overall, our system is designed to enable and test the design and development of ADAS and Human-Machine interfaces for traditional and automated driving scenarios.

Thanks to the latest addition to our simulator fleet, at Virtual Vehicle we are developing innovative concepts and projects that will support the new driver’s roles. We are designing multi-sensory fluid interfaces, capable of tutoring novice drivers of an automated vehicle, seamlessly supporting the transition between automation levels in a comfortable and safe way, according to human-centered design principles. We are also contributing to an ongoing European-funded project (DOMUS) that aims at developing an integrated acoustic and thermal comfort model for drivers of electric and automated vehicles. With our human factors expertise and simulation facilities we are measuring the impact of different cognitive aspects on the subjective impression of sound quality in the vehicle.