This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Tackling car sickness in autonomous vehicles

Author : Phil Morse is a technical liaison at Ansible Motion

07 April 2017

Autonomous driving raises the possibility of more people experiencing car sickness. Fortunately, automotive OEMs have a new tool to tackle this and it comes from an unlikely source, Phil Morse technical liaison at Ansible Motion explains.

A recent study by the University of Michigan’s Transport Research Institute set out to find what people will do in cars once they become superfluous to the driving experience. Researchers Michael Sivak and Brandon Schoettle interviewed more than 3,200 adults on four different continents about what they would do. 

The figures varied considerably, but on average 35.6 percent concluded they would end up taking part in activities that are expected to increase the risk of motion sickness, such as reading, working or watching videos. 

Overall, the report estimated that between 4 and 17 percent of adults – depending on region – would experience ‘moderate or severe motion sickness at some time’ as a direct result of the things they would do in an autonomous car. Another report suggests that as many as 66 percent of adults could experience mild to moderate motion sickness.

That’s quite an alarming statistic for car manufacturers, so naturally the goal is to find ways of mitigating those risks without detracting from other aspects of the in-car experience. Before that can be done, however, there needs to be a way of scientifically studying the onset of car sickness. And that, in short, means making people feel sick on demand.

Fortunately, there’s already a well-proven way of doing this. For nearly a half century, simulators have been used as a means of training aircraft pilots. In more recent decades they’ve been adopted for coaching racing drivers and providing virtual testing for road cars. Over the years, both automotive and flight simulator users have widely reported problems with dizziness, nausea and disorientation. It even has a name: Simulator Adaptation Syndrome (SAS).

The causes of SAS are thought to be essentially the same as those of conventional motion. In both cases, there is a disconnect between the driver’s perception of movement and their other senses. When travelling as a passenger in a car it tends to be due to focusing on things that don’t relate to the vehicle’s motion; in a simulator it’s more often due to a latency between a driver’s inputs and the simulator machinery’s reactions that wouldn’t occur in real-life. The effect, however, is very much the same.

The prevalence of motion sickness in simulators means that the efforts of the engineering firms that build these machines are typically directed towards eliminating it. After all, it is quite difficult to conduct meaningful experiments of any duration if the participants are experiencing discomfort due to the simulator itself. However, eliminating SAS when you don’t want it, in order to prevent false-positives is far from straight-forward. Simulator specialist Ansible Motion claims to have cracked this by taking a clean sheet approach to the design of its Delta series of Driver-in-the-Loop (DIL) simulators. 

Part of this revolves around the use of sophisticated hardware and software to reduce the motion, optical, and other sensory latencies to the point where feedback delays are imperceptible even to professional test drivers. When a driver takes an action, the visceral experience must be genuine, realistic. Achieving this in terms of physics calculations and graphics generation and so on is no mean feat, but achieving the required fidelity from a motion control system is an even greater challenge.

The main reason for this is that the human vestibular system, which provides our sense of balance and spatial orientation, is deceptively complex and inherently non-linear in its behaviour. Simply attempting to replicate or scale down real-world forces doesn’t work within the confines of a laboratory environment. Instead, Ansible Motion uses a radically different approach, by mathematically modelling the vestibular system’s responses as an integral part of its machine control techniques. 

Again, it’s not just a software and control issue. The simulator’s motion system has been designed from the outset to align itself with all six of a vehicle’s primary movement axes. It is also designed to provide sufficiently low inertia and sufficiently high mechanical stiffness to accurately carry out the desired motions in the time required. That might sound like a given, but traditional automotive driving simulator technology has generally been carried over from aircraft applications, where massive, slower-moving “hexapod” pointing machines are perfectly acceptable. The dynamics of road vehicles are characterised by short, sharp movements and trajectory responses, as opposed to the lower frequency movements of aircraft. 

To provide the required agility, Ansible Motion has developed a unique ‘stratiform’ motion system. This simplifies the actuation requirements by placing the driving cabin on top of a series of layers. The first X-Y stage provides lateral and longitudinal movement, while the stages above generate the yaw, pitch, roll and Z-plane ‘bounce’ motions. This layer stack results in a much lower centre of gravity than traditional hexapod machines, which means the control are smaller and easier to manage and manipulate. 

Ansible Motion’s level of control authority and fidelity means that SAS is successfully eliminated for most drivers and experiments. But it also means that motion sickness can actually be purposefully induced within the simulator. This counter-intuitive approach can be used to replicate and study the circumstances that would trigger motion sickness in normal car use.

Several OEMs are now looking into how this can be applied to study the effects of car sickness in autonomous cars. As with other types of simulation-driven experimentation, the beauty of using a DIL simulator for these riding comfort studies is that vehicle engineers are granted with complete and repeatable control over the entire environment. It allows for hands-on testing with real people to begin much earlier and more often in the vehicle development process, saving time and money. Plus, perhaps most importantly, it’s safe.

All this means that DIL simulator testing promises to be a powerful tool for understanding and reducing the occurrences of motion sickness in autonomous cars. And that brings us back to the original question: How might you choose to spend your time in a car that can drive itself?


Contact Details and Archive...

Print this page | E-mail this page

Coda Systems