Ian Haigh, team lead at Ansible Motion, discusses how open access to driver-in-the-loop (DIL) simulation could help more of the automotive industry confront its biggest challenges.
The launch of the open-access driving simulator at Hungary’s Bay Zoltán Research Centre earlier this year marked more than just the arrival of another test facility. Located close to the Austrian, Slovenian and Croatian borders, it’s in a prime position to serve the European automotive industry, including Hungary’s own rapidly developing R&D sector. But what’s special about the new ZalaZone facility is that it grants organizations right across the automotive spectrum access to a crucial development technology that once only existed behind closed doors. The ultimate target is said to be to have all German premium manufacturers develop their vehicles in Hungary by 2030.
Over the last couple of decades, DIL simulation has expanded from a curiosity embraced by a handful of major OEMs and big-budget race teams into an everyday tool for the mainstream automotive industry. Globally, the value of the driving simulator market is predicted to grow to US$2.1bn by 2025.
These days, the applications of DIL simulators range from ride and handling assessments through to HMI and UI development, NVH evaluations and even work on the user experience for autonomous vehicles.
There are fundamental benefits to studying any of these aspects in a simulator, but the unique pressures facing the automotive industry at the moment are driving specific demand in a number of key areas.
Electrification is posing new and sometimes difficult questions. Perhaps most obviously, it brings a whole new character to the powertrain. Driveability and performance feel can be very different when the power, torque and weight characteristics of the vehicles have shifted so dramatically from their combustion-engined predecessors.
Should a high-performance EV, for instance, unleash its full torque capability as soon as there’s enough traction to support that or should it be calibrated to mimic the more progressive delivery of an IC engine? Should EVs be silent, or should they offer some sort of soundtrack? Could the use of torque vectoring or rear-wheel steering mask the extra weight of a BEV powertrain?
The thing that unites all these questions is that they relate to the subjective experience of driving the vehicle. That’s not generally something that can be captured with numerical data from an offline simulation. A car that’s quicker through a simulated lap of a handling circuit, for instance, is not always as engaging or as exploitable as one that might be slower against the clock.
Ironically, this feedback loop that exists between the car and its driver becomes even more critical as the software begins to play a bigger role in the control of the vehicle. ADAS systems, including partial autonomy, are the other big technological battleground in the automotive industry at present.
A key question with any ADAS system is when to intervene, and to what extent. In the EU, emergency lane keeping systems (ELKS) are already mandatory for newly homologated cars and vans. Historically, many of these systems have proved notoriously intrusive, with drivers actively steering against unwanted adjustments or simply disabling the function completely. Again, it’s a matter of human perception.
The stakes can be even higher if an ADAS system is operating under highly dynamic conditions. Imagine, for instance, the ESC calibration on a high-performance vehicle that allows a degree of slip before it intervenes; if the intervention was too pronounced it could even lead to the driver instinctively trying to correct it, making the situation worse rather than better. In this scenario, a realistic and immersive representation of the vehicle’s dynamics is essential for developing and calibrating the system.
In theory, the ultimate test for these systems comes with a physical prototype in the real world, but there are numerous obstacles to consider – particularly in the early stages of development. Physical prototypes can be time consuming and eye-wateringly expensive to produce. Once they do arrive on the test fleet, the competition between different departments to get access to the latest spec vehicles can be intense.
Next comes the job of finding a suitable test venue. Even large OEMs that have their own proving grounds will generally need to ship vehicles around the world for things like winter testing. Once on location, there’s the small matter of finding the right weather conditions and hoping they’ll hold for the duration of the trip. Even small changes in ambient conditions or surface friction can impact the repeatability of the test.
A state-of-the-art DIL simulator such as the Ansible Motion Delta S3 machine now installed at Bay Zoltán is designed to combine the fidelity required to represent these minute changes with total control over the test conditions. If you want to vary the coefficient of friction of the track surface by 1% every lap, for instance, you can do so. What’s more, the model for an initial proof of concept can be ready in days rather than months, and at a tiny fraction of the cost of a physical prototype.
Add up the additional prototype costs, the stoppages due to bad weather or broken cars and the delays to the program before physical prototypes become available, and the return on investment for a DIL simulator becomes clear. What’s harder to absorb is the upfront cost, which can be a seven-figure commitment.
For smaller organizations, and those who only require DIL facilities periodically, this investment can be prohibitive. Under those circumstances it makes sense to do what the industry often does for other pieces of high-capital equipment – rent it from someone else.
Fortunately, simulators such as the Delta S3’s open and modular architecture is software agnostic, making it compatible with all major simulation tools. A model that has been developed and used for offline testing in an industry standard tool such as AVL VSM, IPG CarMaker or Mechanical Simulation CarSim can be plugged virtually straight into the simulator. The center even allows customers to build their own bespoke cabins to use on the shared motion platform and store on-site.
Access to facilities such as Bay Zoltán will allow technology developers to harness the power of DIL simulation. It provides a platform to explore new possibilities earlier in the program and at lower cost than physical testing. And, most importantly, it delivers insights that you can only get with a human being sitting at the controls.
