The need to reduce emissions, provide a more connected experience and increasing levels of autonomous functionality mean that engineers need to do more of their work in simulators rather than on physical prototypes. And it’s the three pillars of electrification, connectivity and autonomous driving that are pushing one of the biggest trends in simulation – the need to conduct a greater number of driver-in-the-loop tests.
Simulation is no longer the domain of combustion engine dynamics and ride and handling. Electrification, connectivity and autonomous functions create a new range of scenarios that engineers haven’t previously had to consider so intensely: differing acceleration profiles, motion sickness, driver handover and infotainment-human interaction are just some of the examples being researched.
And driver-in-the-loop simulation is important because, no matter how many artificial tests you run, they can rarely include the nuances found in the human interaction with the vehicle and will always be an approximate reality. This is heightened even more by the growing complexity of vehicles.
Ten years ago, there might have been 20 to 30 microprocessor-based systems on a car; now it could be 70 to 80 with individual sub-systems running their own code. That means an explosion in the number of possible interactions between the vehicle and its occupants.
Driver-in-the-loop simulators can challenge a driver with 100 once-in-a-decade environments, or challenge 100 drivers with identical scenarios, allowing engineers to develop statistical confidence in the reaction of systems to humans and vice versa.
And, as the number of electrified, connected, autonomous vehicles increases, simulation allows firms to test integrated hardware, whether it’s ultrasonic sensors, cameras or lidar for automated systems, acceleration and deceleration surges in EVs or screens in infotainment systems, making validation simpler.
But technologies such as autonomous functionality arguably demand greater levels of immersivity and require the ability to create scenarios that test the hardware – a camera doesn’t need perfect graphics but does need pedestrians and other cars to behave as naturally as possible – and the humans who interact with it.
While vehicles continue to develop, humans haven’t changed, so there’s a focus on improving how the vehicle communicates with its occupants. A lot of work is being conducted inside driver-in-the-loop simulators on the cabin environments and how they can communicate the driving experience.
There’s always been a requirement for light, low-latency, precise acceleration and accurate motion to stimulate the driver’s vestibular system, and there’s a trend for these systems to grow larger in footprint so that you can sustain the vestibular stimulus for longer periods.
The challenge is the cost and space needed to conduct the simulations. Having one huge system in a company is not a very good answer when that company’s developing 20 models in parallel, with perhaps another 60 derivatives of those models, with every team wanting access to the single simulator.
This is leading firms such as Ansible Motion to develop portable simulators that can be moved between departments as and when necessary to meet demand. Analysts predict that the driving simulator market will grow to reach $2.4bn by 2025.
There’s always going to be a place for physical tests as they will remain the final proof of the pudding before vehicles go into the market. But a physical test is expensive, very hard to repeat precisely and very time consuming – which is where simulation has an advantage.
And, because the industry is moving so quickly towards electrified, connected, autonomous vehicles, it needs a greater understanding of the technologies’ impact on the human, which is why driver-in-the-loop systems will become increasingly important.
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.