Local, foreign researchers warn: Roadway automation will bring on fresh problems

Work done at Stanford, BGU highlights issues we’ll face when we want to take control of autonomous vehicles

By JUDY SIEGEL-ITZKOVICH
Cars on a highway [illustrative] (photo credit: REUTERS)
Cars on a highway [illustrative]
(photo credit: REUTERS)
When human drivers take back control of an autonomous car they will have problems steering, according to researchers at Stanford University in California and Ben-Gurion University of the Negev.
It all depends on how driving conditions changed since they were last at the wheel.
A total of 22 drivers put the matter to a test – on a track, not a freeway. The results, which were just published in the first issue of Science Robotics, could help in the future design of autonomous cars, said Lene Harbott, co-author of the research and a research associate at Stanford.
There you are, cruising down the freeway, listening to some tunes and enjoying the view as your autonomous car zips and swerves through traffic. Then the fun ends and it becomes time to take over the wheel. How smooth will that transition be? The researchers – who have expertise in autonomous car design, human-robot interaction research and neuroscience – found that the transition could be difficult. Drivers who experienced certain changes in driving conditions since their last time at the wheel, such as changes in speed, experienced a period of adjustment in their steering.
“Many people have been doing research on paying attention and situation awareness. That’s very important,” said Holly Russell, lead author of the research and former graduate student in Stanford’s Dynamic Design Lab.
“But, in addition, there is physical change, and we need to acknowledge that people’s performance might not be at its peak if they haven’t actively been participating in the driving.”
IIana Nisky, co-author of the study and senior lecturer at BGU, said, “In neuroscience, this is explained as a difference between explicit and implicit learning. Even when a person is aware of a change, their implicit motor control is unaware of what that change means and can only figure out how to react through experience.”
Nisky is head of the Biomedical Robotics Lab and a member of BGU’s biomedical engineering department.
The trouble adjusting to different driving conditions was not enough to cause drivers to miss their turns, but it was enough to be noticeable in how they wobbled the wheel to account for over and under steering. These challenges, the researchers said, raise the possibility that – depending on the particulars of the driver, driving conditions and the autonomous system being used – the transition back to driver-controlled driving could be an especially risky time.
Study participants drove a 15-second course consisting of a straightaway and a lane change.
They then took their hands off the wheel and the car took over, bringing them back to the start. After doing that four times, they drove the course 10 more times, with steering conditions that modified to represent changes in speed or steering that may occur while the car drives itself.
Changing the steering ratio from the standard 15:1 to 2:1 simulated the more sensitive steering that drivers experience at higher speeds. This modification made the car turn more sharply, to simulate the way less steering wheel movement is needed to make a lane change at a high speed, as opposed to that needed a low speed.
All drivers were given advance warning of the changes and had some opportunity to adjust to the difference during the straightaway. Regardless, during the altered steering ratio trials, the drivers’ steering differed significantly from that demonstrated prior to the modifications.
“Even knowing about the change, being able to make a plan and do some explicit motor planning for how to compensate, you still saw a very different steering behavior and compromised performance,” said Harbott.
After being driven back to the start by the car, with the original conditions restored, participants drove the course six more times. Drivers who experienced the steering ratio change again displayed a clear period of adjustment, undershooting the steering needed to complete their lane change.
This driving test is close to a classic neuroscience experiment that assesses motor adaptation.
In one example of these experiments, participants use a hand control to move a cursor on a screen. The way the cursor moves is adjusted during the experiment, and participants, in turn, change their movements to make the cursor go to where they want.
Just as in the driving test, participants in the experiment have to adjust to changes in how the controller moves the cursor. They also must adjust a second time if the original conditions are restored.
“Even though there are really substantial differences between these classic experiments and the car trials, you can see this basic phenomena of adaptation and the after effect of adaptation,” said Nisky. “What we learn in the laboratory studies of adaptation in neuroscience actually extends to real life.”
It also showed that the effect and after-effect of motor adaptation applies to skilled tasks learned over a long period of time.
Although the drivers were not so thrown by the changes in steering that they drove offcourse, the period of altered steering behavior is still significant, she continued.
“There are so many different variables involved in driving that anything that compromises driving performance could lead to an accident.”
The test vehicle was developed at Stanford for the study and does not represent any currently available system. The study addressed one specific example of handover, but there is still a lot to learn about how drivers respond in other circumstances, depending on the type of car, the driver and how driving conditions have changed, researchers said.
“If someone is designing a method for automated vehicle handover, there will need to be detailed research on that specific method,” concluded Harbott. “This study is the tip of the iceberg.”


Related Tags
design