California bike club helps Aurora self-driving car to spot motorcycles
It’s a known fact that motorcyclists are at great risk for accidents and injuries from motor vehicles because of the lack of protection afforded. One of the most common causes of personal injury on a bike is when a motor vehicle driver doesn’t see it. If self-driving cars become a popular choice for future drivers, will they become a help or a nightmare when sharing the road with cyclists?
Enter the San Francisco Chapter of the Iron Order Motorcycle Club who sent six of their members to ride and maneuver around a self-driving test car, using Aurora’s computer technology for an event dubbed “Motorcycle Learning Day.” Aurora is a company headed by a former Tesla tester that is developing an autonomous driving system (not a car) named Aurora Driver. In addition to testing the computer-controlled reaction of their car to regular motor vehicles (both on the road and simulation), it also tests errant shopping carts, construction crews, e-scooters, bicycles and pedestrians.
Aurora asked the club to help because the prototype is not as used to driving around motorcycles as it is driving around more easily-seen vehicles on the road. The car and its developers needed to see how to make the Aurora Driver sensitive to cyclists, both in computer-driven or manually-driven modes.
The day-long testing called for the car to be placed in manual mode with a driver, to learn how to detect and recognize a motorcycle, collecting data as it drove. The riders drove a variety of motorcycles and performed regular traffic activities around the test car. This included such actions as lane splitting (allowed in California), merging in front and behind it, and riding alongside the car in different positions.
A very important capacity for an autonomous car is its ability to brake and avoid a rear-ender. Aurora understands that it’s one thing for a self-driving car to learn to brake when there is a motor vehicle in front of it, and quite another thing when the object in front of it is a motorcycle. The question was whether the car could intuitively recognize a motorcycle if it pulled up to the edge of the lane. If not, it has to be “taught” to do so.
Aurora says that ultimately, the self-driving car will be able to recognize various models and types of motorcycles, identifying the difference between a cruise and a sport bike. Discerning the difference is important when predicting and managing closing in speeds or merging patterns.
The data collected during Motorcycle Learning Day was just one of many learning situations for collecting data for Aurora’s computerized driving system. It is especially important in light of a report that came out in August compiled by the European industry group ACEM. Their published findings said that among the “vulnerable road users” such as pedestrians, bicyclists, and others, “Motorcyclists are the only road users who share all kinds of road and traffic environment conditions, including full velocity range, with (manually-driven) cars. This creates a major safety challenge.”
They went on to say that the technology they studied was good at identifying large objects, but not very good at detecting smaller objects, like motorcycles. Their report cited a 2016 Netherlands’ study, reporting that some adaptive cruise control systems failed to detect and match speed with motorcycles in front of them. If that is the case with human-driven cars, what’s to happen with self-driven cars?
Some experts hope that whatever is learned from the Aurora testing will also help develop better sensors on regular cars.
Sources: Mashable and Ride Apart
If you have any questions about this topic or any other questions related to personal injury law, please call us at 920-725-8464, or toll free at 1-800-529-1552. Our personal injury consultations are always free.
Please click on https://www.direnzolaw.com/blogto read our other personal injury articles.
The content of this blog was prepared by Law Offices of DiRenzo & Bomier, LLC for educational and informational purposes only. It is not intended to solicit business or provide legal advice. Laws differ by jurisdiction, and the information in this blog may not apply to you. You should seek the assistance of an attorney licensed to practice in your state before taking any action. Using this blog site does not create an attorney-client relationship between you and Law Offices of DiRenzo & Bomier, LLC -client relationships can only be created by written contract.