It is a matter of time before self-driving cars are coasting down our streets on a daily basis. As companies work tirelessly to produce self-driving cars, they work on getting the cars to recognize pedestrians, birds, and so forth.
However, self-driving cars are struggling to identify bicycles, posing as a threat to the pedestrians that use bicycles as their main form of transportation.
What is the Main Issue?
According to Nuno Vasconcelos, a visual computing expert at the University of California, “A car is basically a big block of stuff. A bicycle has much less mass and there can be more variation in appearance.” Even if self-driving cars do see bicycles, they don’t know how to identify them.
Can Self-Driving Cars See Bikes?
Self-driving cars are struggling to see bicycles. Researchers at George Mason University found that when they worked to get self-driving car technology to recognize objects, the technology could identify 89 percent of cars. However, it could not manage to spot and identify many bikes or bicyclists. In fact, it only recognized 74 percent of bicycles.
What Can Be Done?
Essentially, self-driving car producers have to work harder to get these new technologies to identify bicycles and the people riding them. For instance, Google’s autonomous vehicle spinoff has received upgraded bike-recognition capabilities, tackling this issue head-on. Ford claimed that “highly detailed 3D maps” are going to be at the center of the self-driving test cars they are putting out this year.
Are Bicyclists in Danger?
The short answer: no. Bicycles are fast and it can be hard to tell what they might do next, making them relatively unpredictable. Self-driving cars are struggling to see them but that is something that can be worked on through improved technology. Human error alone results in most traffic fatalities, so self-driving cars aren’t putting bicyclists in any more danger than they are riding down the street on any given day.