Google introduced the self-driving car project at 2010, but this innovative project gets smarter and efficient every day. These days, Google announced that over the last year they’ve been working almost entirely on mastering city street driving.
Naturally, driving in the city is a lot different than driving in the freeways, as constant distractions can appear, like car not stopping at a red light, kid jumping to the street, cars double-parking and blocking lanes, and much more. These situations can be difficult to handle even by experienced drivers, but not so much to the computer operating the car.
So far, the self-driving car logged thousands of miles on the streets of Mountain View, CA. The car’s software can detect hundreds of distinct objects simultaneously – buses, pedestrians, a stop sign held up by a crossing guard, or even a cyclist indicating a possible turn.
Situations that seem chaotic and random to the human eye are actually quite predictable to a computer. Countless scenarios were loaded to the software models, from likely situations such as a car stopping at a red light, to unlikely situations such as a car blowing through the red light. There are still many problems to solve, but thousands of city driving situations can now be navigated autonomously.
Google’s self-driving car have logged 700,000 autonomous miles so far, and the goal of a vehicle that operates fully without human intervention is now a lot more achievable than ever.