In the past few years, Google has been testing its driverless car on highways in several states, including Nevada, California, and Florida, logging about 700,000 miles without any reported accident, but even though it’s a very strong indication of how reliable and safe Google’s technology is, it is not enough to convince people that autonomous cars are ready for commercial use, given that it is yet to be tested in urban areas, which is much more complicated and difficult than driving on the highway.
That’s why the tech giant recently started a project for preparing its cars for city driving, by improving the software that helps the vehicles detect and recognize various that they could encounter while driving in an urban setting. Google’s self-driving car is already equipped with a series of radars, cameras, lasers, and GPS systems, which allow it to maintain a safe following distance behind other vehicles, stay in its lane, and get to a predetermined destination on its own, but driving on city roads brings many additional challenges, with lots of unique and delicate traffic situations that are often difficult to deal with.
There are plenty of different objects that cars have to recognize when driving on city roads, such as pedestrians, cyclists, traffic signs and signals, so Google had to upgrade its technology and equip the cars with software that will allow them to do that, given that the existing technology is not capable of that. Google says that the improvements that it made to the software make it possible for its self-driving cars to recognize traffic cones, follow cone layouts and change lanes, as well as maintain a safe distance from various obstacles. Furthermore, they should be able to recognize gestures that cyclists make to let motorists know that they are about to make a turn, which is a pretty amazing achievement, and one that can be very helpful in reducing the number of bicyclist fatalities, given that most of them occur in urban areas as result of collisions between cyclists and motor vehicles.
Since many accidents occur when cars approach near construction zones, Google has been trying to find a solution that would make these situations less dangerous. The company says that their sensors can now recognize the yellow signs and cones early, and inform the car’s on-board computer that it has to change lanes. Another great thing about the new and improved technology is that it can help the car detect and track objects behind it.
However, no matter how sophisticated and superior this technology might seem, there are still a lot of issues that Google and other companies that are developing self-driving have to address before they can become commercially available. At the moment, one of the biggest obstacles is the high cost of autonomous driving technology, along with legal issues related to accidents involving self-driving cars. In any case, automakers and tech companies are confident that these problems will be resolved in the next couple of years, and hope that these vehicles will be able to hit the road by 2025.