A few issues plaguing Google’s self-driving car

Mar 3, 2013
11

Self-driving cars are the future and various groups are currently working on perfecting their self-driving cars and preparing them for the masses. Oxford University tested out their own self-driving car system recently to much success. It's able to perform well on regular commutes and in traffic jams, and the team is currently working on a system that will allow their car to easily map surrounding areas. Google's self-driving car system is doing well too, being able to perform daily tasks with ease, and most of time being able to drive better than humans. A few reports from Google's team, however, state that there are a few flaws that need to be worked on.

While Google's self-driving car may be able to handle normal driving tasks, it can hit a few roadblocks under certain circumstances. These problems are fixable, but will require much creative thinking amongst the Google team (something they are obviously pretty good at). The 3 issues plaguing the self-driving car system include: Driving in snow, driving in areas with unsuspected changes in the road, and driving in areas where a human is directing traffic.

While driving in the snow, the car's sensors have a hard time analyzing the lane markers on the road, which it needs to be able to do in order to stay correctly in its lane. When the cars come across an unexpected change in the roads that is not recorded in its mapping system, it becomes lost and is unable to find its way around, or find an alternate route to its destination. And last, while driving through areas where traffic needs to be directed by a human, such as construction zones, the car is unable to analyze the traffic director's gestures and thus becomes confused.

The engineers are currently working on fixes for these problems, and it may require some out-of-the-box thinking. Development of self-driving cars is advancing further and further and I can't wait for them be available to the masses. Currently, the prices of self-driving car systems are speculated to cost an arm an a leg, but the teams are working on reducing those costs down to a much more affordable price tag.

[via Business Insider]


Must Read Bits & Bytes