It’s pretty clear already the future is all about electric vehicles and self-driving cars, so it’s not necessarily a surprise that more and more companies are trying to make sure they have a well-deserved place in this new side of the automotive industry using revolutionary ideas.
This time, it’s Amazon’s turn to come up with a technology that would actually make sense on pretty much any self-driving car out there.
A patent called “Systems and methods for utilizing images to determine the position and orientation of a vehicle” describes a newly-invented technology that Amazon believes could help address a major shortcoming: the reliance on GPS for autonomous vehicles.
“GPS can be inaccurate, irrelevant, unreliable, unavailable, etc. in certain environments,” Amazon explains in the patent. “Moreover, GPS may have a degree of accuracy or tolerance that may be insufficient for certain situations.”
So basically, Amazon doesn’t think using GPS alone is the right way to go for a self-driving car that always needs to know where it is, especially given these problems. So the company has invented a way to deal with the whole thing.
The patent describes multiple ways to determine the location of a self-driving vehicle, even in areas without a GPS signal. One method, for example, involves taking photos of the surroundings and then comparing them to a reference map, helping the car understand exactly where it is.
If this is possible, the system could prove even more accurate than a common GPS-based implementation, as the location would be determined with a precision of down to just a few meters, all by simply comparing a set of photos with those in an existing database.
But why is Amazon investing in such systems? The answer is as simple as it could be.
Amazon is one of the companies that are betting big on self-driving cars in the future, not because it wants to be involved in the automotive industry but as part of its shipping services. Amazon wants self-driving cars to ship packages virtually everywhere, so it’s pretty clear that spotty GPS connections are a major shortcoming.
Of course, this system is a pretty good idea, but let’s not forget it’s still in the patent stage. It could take years until it reaches mass production, but given the world is not yet ready for self-driving cars, the timing seems just about right.
A patent called “Systems and methods for utilizing images to determine the position and orientation of a vehicle” describes a newly-invented technology that Amazon believes could help address a major shortcoming: the reliance on GPS for autonomous vehicles.
“GPS can be inaccurate, irrelevant, unreliable, unavailable, etc. in certain environments,” Amazon explains in the patent. “Moreover, GPS may have a degree of accuracy or tolerance that may be insufficient for certain situations.”
So basically, Amazon doesn’t think using GPS alone is the right way to go for a self-driving car that always needs to know where it is, especially given these problems. So the company has invented a way to deal with the whole thing.
The patent describes multiple ways to determine the location of a self-driving vehicle, even in areas without a GPS signal. One method, for example, involves taking photos of the surroundings and then comparing them to a reference map, helping the car understand exactly where it is.
If this is possible, the system could prove even more accurate than a common GPS-based implementation, as the location would be determined with a precision of down to just a few meters, all by simply comparing a set of photos with those in an existing database.
But why is Amazon investing in such systems? The answer is as simple as it could be.
Amazon is one of the companies that are betting big on self-driving cars in the future, not because it wants to be involved in the automotive industry but as part of its shipping services. Amazon wants self-driving cars to ship packages virtually everywhere, so it’s pretty clear that spotty GPS connections are a major shortcoming.
Of course, this system is a pretty good idea, but let’s not forget it’s still in the patent stage. It could take years until it reaches mass production, but given the world is not yet ready for self-driving cars, the timing seems just about right.