Google (NASDAQ:GOOG) started out its journey as a online search provider and over the years has ventured into innumerable businesses. But this doesn’t mean Google has reached all that it can be. Each day the company adds hundreds of productive hours behind making the impossible possible. And now the company is thinking of taking the next big leap into the world of virtual reality, where it wants to make self-driving cars that will run by making the real world more understandable to robots.
It would be an understatement if we say the technology in play is complicated. It’s beyond complicated and probably possible only for a company such as Google that already has the expertise and experience in gathering the required kind of data. Frankly speaking it’s all that Google has earned in all these years that’s driving its growth. Let me explain to you how the technology works and where the difficulties are.
Technology in Play
Most of the consumers will not be interested to know how the driverless cars from Google work as long as they know they are safe inside it and can talk and text as much as they want to. But for analysts, industry experts and people with the craving for the knowledge, the technology under the hood is of immense interest. These cars don’t go on scanning and reading the roads of a city in real-time while travelling. Instead they are getting all kind of assistance from Google.
- Warning! GuruFocus has detected 5 Warning Signs with GOOG. Click here to check it out.
- GOOG 15-Year Financial Data
- The intrinsic value of GOOG
- Peter Lynch Chart of GOOG
What Google is doing is it’s sending out teams out in the real world to gather as much relevant data as they can and then the data is being fed into a system which acts as the brain of the driverless cars. Google has task forces that travel each road and map every minute detail and that also with great preciseness so that the cars can know which curb is how high and where it can and where it can’t ride. It’s like Google is trying to capture the real work and transform it into a virtual world that its robot cars can better understand and read.
The company is investing huge amount of money and muscle power to keep this project running. The technicians provides all the information required, such as the height of the traffic signals from the ground, what’s the speed limit for the driving zone, which direction the traffic is supposed to flow, possible mishaps and road blocks information available live from satellite transmissions. Putting it simply, all possible circumstances are readily available to the cars and all they have to do is check if the actual situation is similar to what’s in their records. If yes, they can very well roam around the place, and if no, they can analyze the situation using the standard equipments (the radar, laser sensor and the cameras) and then proceed.
Barriers to Implementation
Surely the technology looks interesting and make the life of the rider very easy. But putting the technology together is not an easy task and makes the life of the tech guys very, very difficult. After years of hard work Google has been able to map and create virtual image of Mountain View only and its cars can freely roam around the place without any difficulty. But if someone thinks of driving the car in New York or Orlando, he or she will be greatly disappointed. Outside Mountain View, the cars will fail to demonstrate their capabilities.
The biggest obstacle in commercial implementation of the cars is the lack of detailed data. Surely Google has a bulk of data stored in its servers thanks to Google Maps, but the data is not detailed enough or ready enough to support the driverless cars. These robot cars don’t rely on just map data, instead they need extremely detailed layout of the real world, so that they know what to expect.
So far Google has been able to map only 2,000 miles of US roads and 99.95% of US roads are yet to be mapped. That’s a huge work that is still ahead of Google. The only respite the company has is that they have the expertise and the experience to carry out such a hefty activity. Again, capturing the data just once will not be enough to support the movement of the cars. If the description of a place changes, say a new monument is created or a highway has been constructed, Google will have to map that promptly and make all the necessary changes.
Despite huge headwinds, Google is pretty determined to get the robot cars out in the market and make it ready for mass adoption. Surely the company doesn’t expect to do this in the immediate future and understands this is going to take a fairly long period of time. What Google is banking on is that the driverless cars are smart machines equipped with state of the art gadgets. Even if Google is not able to map a geographic area, if a buyer uses the car in that area, he will himself be able to feed in the data into the car’s system by manually driving the car a few times to the destination. Once the system absorbs the data, the user can sit back and relax and enjoy a stress free ride back home, probably listening peacefully to his favorite music.