Self-driving Lexus Hybrid

Google’s Self-Driving Cars Tackle The Urban Jungle With Surprising Ease

Long before Nissan, Tesla, GM, Mercedes-Benz, Renault, or Ford promised us self-driving cars, software giant Google was giving closed-road demonstrations of its own self-driving technology. It even famously once took a visually-impaired man on a trip around his neighbourhood to carry out some essential chores, albeit accompanied by the local highway patrol and film crews.

Google's self-driving car program has now graduated to a bigger car -- and bigger streets.

Google’s self-driving car program has now graduated to a bigger car — and bigger streets.

Since we last saw Google’s self-driving cars however, they’ve graduated to the next stage: real-world tests in and around the Mountain View, California base of Google.

Google’s latest update on the self-driving car project is impressive. Not only have its self-driving cars — now Lexus RX 450h hybrids instead of the Toyota Prius originally used in the project — completed countless miles of real-world, accident-free driving in all kinds of weather and traffic patterns, but Google’s self-driving software has got smarter too.

As a YouTube video accompanying Google’s update shows, its self-driving hybrids can now spot construction zones and parked cars, safely manoeuvring out of the way to avoid any collisions and autonomously following any temporary lane changes.

In addition, the vehicle can now act appropriately at railroad crossings, ensuring that the car does not pass through the crossing until it knows the exit is clear, while a special cyclist detection subroutine can predict and react to the movements of pedestrians and cyclists.

Impressively, Google’s software can now recognise the hand signals of cyclists, giving the cars forewarning if the cyclist intends to turn. What’s more, the software treats cyclists with caution, behaving in a proactive way to enable the car to predict the typically unpredictable behaviour of humans on two wheels.

What’s more, the system can even detect a cyclist or motorcyclist as they filter at a stop light, ensuring the car doesn’t turn when there’s someone or something in what would traditionally be a driver blind spot.

Google's self-driving system operates differently to Nissan's autonomous drive system

Google’s self-driving system operates differently to Nissan’s autonomous drive system

Google says it’s a long way from completing its project, and admits there are many more hours of driving to put in before its self-driving cars can autonomously navigate each and every street in Mountain View without any human input. But like the other self-driving cars we’ve covered recently, it shows us that autonomous cars are far, far closer to reality than we ever thought before.

We’re curious to see how long it will be before Google uses its self-driving car in all-electric cars, but regardless of the fuel type, autonomous driving technology can save fuel and reduce accidents. And that can only be a good thing.

Would you trust a self-driving car? Have you seen the Google self-driving SUVs driving around the San Francisco Bay area? Leave your thoughts and sightings in the Comments below.


Want to keep up with the latest EV news? Don’t forget to follow Transport Evolved  on Twitter, like us on Facebook and G+, and subscribe to our YouTube channel.


Want to keep up with the latest news in evolving transport? Don’t forget to follow Transport Evolved on Twitter, like us on Facebook and G+, and subscribe to our YouTube channel.

You can also support us directly as a monthly supporting member by visiting

Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedInDigg thisShare on RedditEmail this to someonePin on Pinterest

Related News

  • D. Harrower

    This is fascinating to watch develop! I’m curious, does the vehicle have to be specifically programmed for each locale, or do the new avoidance algorithms theoretically work anywhere?nnnMy folks own a Lexus RX400h.

  • Surya

    I’m sure that by the time this technology comes to market, we shouldn’t fear a thing. It will be much better than an error prone human. They wouldn’t never approve it for use if it wasn’t.

Content Copyright (c) 2016 Transport Evolved LLC