You’ve probably been hearing a lot of buzz about self driving cars and may have also heard conflicting stories about their availability. In 2015, Google was predicting that an automated Tesla would be available by 2018. On the other hand, automobile realists claim that we won’t be seeing them regularly driven on the roads until much later.
A fully autonomous vehicle may still be years away as the software continues to develop to keep up with the hype. Testing has shown that their reliability to avoid accidents still needs much improvement and these driverless cars should be considered to be a scientific experiment. With any experiment of this kind, you don’t know what the answer will be until you get there.
Deep learning has certainly come a long way during the last decade in the tech and AI industry. Perhaps this is why the car manufacturers have become so optimistic about the timeline. It has been used in a number of different applications including the news feed on Facebook and for Google Search, but a large amount of data is required to make it work properly. Every type of scenario that could be encountered must be worked into the algorithm. Using generalizations hasn’t worked up to this point.
You can just imagine how many different scenarios would have to be incorporated into the algorithm before the cars can be put to market. It’s difficult to think that all scenarios could even be covered since there are so many different things that can come up on the roads. Right now it’s hard to predict the measure of success that engineers will have when it comes to isolating the scenarios and handling potential errors.
When will fully automated cars become a mainstay on our roadways? Well, time will tell. In the meantime, if you have any questions or concerns about self driving vehicles, questions about the services we offer, or would like information about fleet management, please visit our website at www.qodelogix.com.