The freedom to innovate

Getting back to engineering fundamentals with driver-in-the-loop simulation

Just suppose for a second that Frank Whittle had never been able to tinker with his earliest ideas about the turbojet engine. Or that Tim Berners-Lee had run out of time before developing what went on to become the World Wide Web.

History is surely littered with great inventions that never happened. Of course, we’ll never know what they were, but it’s fair to assume there have been thousands of them. Imagine all the bright ideas that might have been, were it not for time pressures, budgetary limits, technological constraints or logistical issues.

Right now, there are few sectors more turbulent than the automotive industry. The full-scale charge towards electrification (no pun intended) has forced manufacturers to re-evaluate almost every area of the automobile. From the vehicle dynamics impact of heavier powertrains and low-slung battery masses, through to the UI/UX challenges of helping the drivers manage their energy usage, almost every aspect of modern vehicle design changes in some way with an EV.

And that’s by no means the only challenge facing the automotive industry at the moment. Increasingly sophisticated ADAS, self-driving functions, greater demands for in-car connectivity and the move towards new business models, such as subscription services and shared mobility, are all vying for design resources too. Plus, the pressure is always on to shorten development cycles and reduce costs.

Although all this disruption has the potential to prompt innovation, it can also stifle it. With the stakes being so high, is it really worth gambling on the development of a novel solution when a well-proven one will do? What is the value proposition of introducing new technologies in the context of maintaining and elevating brand identity? These are the difficult questions facing automotive manufacturers worldwide.

Fortunately, cutting edge research and simulation technologies such as Driver-in-the-loop (DIL) simulation allow vehicle development engineers to shift the odds dramatically in their favour. With DIL simulation, for instance, a virtual prototype for proof of concept investigations can be put together in a matter of days at minimal cost, whereas a physical prototype can take six months and more than half a million pounds to build. Multiply that out across an entire prototype fleet, and it becomes easy to see why management might think twice before signing off a high-risk programme unless it can be supported by appropriately advanced tools and techniques.

DIL simulation largely removes the risks associated with these ‘what if?’ questions, while simultaneously expanding the sandbox. Imagine, for instance, that you were planning a new electric vehicle platform and pondering whether the dynamic benefits of running an individual motor for each wheel justified the additional packaging complexity. Or perhaps the outgoing model used MacPherson Strut suspension and you wanted to evaluate whether the cost of re-engineering it for a double wishbone setup would be justified.

DIL simulation provides the freedom to pose these questions, months ahead of any physical builds or testing. It also provides an ongoing benefit with total freedom to vary the test conditions combined with laboratory-calibre repeatability.

There are no logistical issues to worry about, either. With physical testing, there’s always a danger that you’ll arrive to discover that a winter testing  venue is experiencing a warm snap or that a desert proving ground is under flash flood advisory. Even if everything goes to plan, it takes a considerable amount of time and money to ship a fleet of prototype vehicles around the world. The environmental impact of doing so can’t be ignored either.

The same applies to off-line testing to a certain extent, but when it comes to vehicle attribute decisions, there’s no substitute for having a human driver (or occupant) inside the loop, actually interacting with and experiencing the vehicle. Human beings perceive things that might not be immediately apparent in numerical data. For instance, an ADAS system such as Lane Keep Assist may fulfil all its on-paper design criteria, but only a human driver can judge whether it feels too intrusive or too eager to intervene.  

But the benefits of placing a human in the loop go way beyond subjective assessment. When we are in a car, we are experiencing an on-going feedback loop with the vehicle; for example, each vehicle control input – say, a steering, throttle or brake correction – is a carefully calculated response from our brain, based on dozens of different vehicle stimuli combined together and interpreted by senses. Without those uniquely human interpretations and responses, it’s impossible to get truly representative inputs for, say, a test bench or an off-line vehicle simulation model. Fundamentally, the human is a vital part of the equation, even when it’s a matter of collecting purely objective, numerical data.

Returning to the ADAS example: an aggressive intervention from an ostensibly assistive system might inadvertently trigger a human driver to overcorrect, doing more harm than good. Another classic example is high performance optimization related to overall vehicle stability; pick a suspension and steering setup that’s too aggressive in terms of, say, yaw from steer response, and it may theoretically be better, but you risk creating a car that even a Formula 1 driver would struggle to control. It’s a matter of keeping engineering fundamentals in mind while being immersed in a large number of new automotive technologies – as well as a matter of having access to safe and robust exploratory tools such as DIL simulation.

A DIL simulator is a considerable investment, it’s true, but it’s one that pays for itself quite quickly. Continental, for example, has recently invested in an Ansible Motion Delta series S3 simulator, and predicts that it can save around 10,000 tyres and 100,000 test kilometres per year as a result. Savings for vehicle OEMs that currently run large prototype vehicle test fleets could be greater still.

It’s also worth noting that DIL simulators are becoming more accessible than ever. The Bay Zoltán Research Centre in Hungary recently began offering open access to another Delta series S3 simulator. Designed to be software-agnostic, their DIL simulator lab allows engineering teams to bring their own vehicle and environment models, created in virtually any major simulation environment. The Centre can even support customers who wish to build their own bespoke cabins to use on the motion system.

All of this means that it’s now easier than ever for engineers to delve into the important questions that surround modern vehicle design and development. Ultimately, DIL simulation and other cutting-edge tools provide the freedom to explore and innovate, by making use of advanced technologies that are commensurate with the advanced state of contemporary automobiles. In a world that’s all too frequently constrained by budgets or logistics or technology itself, it is nice to know that there are tools that enable engineers to focus on engineering once more.