Research quantifies car sickness for driverless automobiles

Thursday, September 12, 2019 - 6:35pm

.

Design by Michelle Fan

The University of Michigan Transportation Research Institute released a report detailing methods to quantify motion sickness in automobiles in August. With the conversation on driverless vehicles growing, motion sickness has been a concern, since one out of three people is highly susceptible to motion sickness according to the National Institutes of Health.

UMTRI Director James Sayer, an adviser to the project, said the autonomous car industry talks about how these cars would free up time for passengers, allowing them to even watch movies. But motion sickness needs to be addressed before this idea can be made into reality, Sayer said.  

“The idea that you could have a fully autonomous vehicle drive you to work while you get work done or watch a movie … that’s something that the industry has been pedaling for some time,” Sayer said. “We have to start thinking more seriously about why passengers can’t currently watch movies. How is that going to change just because the vehicle is autonomous?”

Monica Jones, an assistant research scientist at UMTRI, is leading the investigation. When digging deeper into published literature on motion sickness, her team found that there was a lot of work done on motion platforms, simulators and other modes of transportation, but not road vehicles. They also found that there was not a way to quantify motion sickness and compare one solution or countermeasure to another.

The study was initially funded by Mcity, where they developed an in-vehicle testbed consisting of a test drive, in-vehicle instrumentation and measurement protocols. 

The scripted test drive squeezed 25 braking events, 45 left turns and 30 right turns into a 20-minute drive. The drive was conducted at low and moderate acceleration, and passengers were given two scenarios. One allowed the passenger to have free head movement; the second restricted the passenger’s head movement by giving them content to read on a mini iPad. 

While the scripted test drive was successful at inducing motion sickness, another concern arose.

“But, by doing that constant hand-turns … it was relentless in terms of the number of maneuvers you would experience in 20 minutes,” Jones said. “The pushback was ‘That that’s not normal. We would never do that many roundabouts in a row.’”

To solve the issue, the team began developing a new protocol to more accurately mimic passenger experiences last summer.

Victor Le, a mechanical engineering doctoral candidate, helped develop the new protocol.

“We carried on the project with the intention of bringing it onto real-time traffic,” Le said. “We wanted to emulate real-time traffic as closely as possible while maintaining and improving the protocol we used in the 20-minute drive in Mcity.”

Today, the test drive has been scaled to 60 minutes, and the team is planning on collecting a larger sample on the routes this coming summer. The test drive is led by a trained driver who has spent hours replicating the route to make sure it is consistent. They hope that the new data will correspond to the data samples collected previously.

“It helps us validate what we developed in Mcity,” Jones said. “If the response in 20 minutes — even though it’s intense — is some field representation of what we see on-road, then it becomes a great testbed for us to look at mitigation factors.”

Looking toward the future, Le sees this project being used when fully autonomous vehicles have been created. 

“I think once autonomous vehicles get to the point where we can run tests with them, we would be able to test autonomous vehicles without real-time traffic while maintaining and measuring metrics we think are most indicative of motion sickness,” Le said.