One of the best demos I’ve seen at Qualcomm 5G Summit is this 5G positioning development which uses a combination of 5G signal and artificial intelligence to enable positioning of a device with an accuracy of 1 foot.
As you may know, consumer-level indoor tracking has been in the works for over a decade. However, they still don’t have access to anything remotely practical like location services in malls.
There are mutliple reasons for this. Most indoor tracking systems require special infrastructure (which nobody wants) and use wireless protocols that weren’t standard yet. A direct line of sight between the transmitter and receiver can be lost, resulting in poor position accuracy. Finally, the signal can pass through objects such as trees that interfere with the signal.
Qualcomm’s approach to 5G positioning is twofold. First, 5G location uses wireless communication infrastructure. It’s already there and not just installed for location services.
It works by sending your phone (or other 5G device) a very simple 5G signal with a time stamp to nearby 5G access points. The amount of data transferred is tiny and consumes virtually no bandwidth.
Each access point receives the signal at a slightly different time, and from there it is possible to triangulate the transmitter’s position. It’s the same principle as the GPS system, but indoors. In itself an ingenious and practical solution.
However, that’s not quite enough. Unfortunately, there are many situations in the real world where walls, corners, trees, etc. can obscure the signal. All of these things can reduce accuracy down to 8 yards, which is too imprecise for many use cases.
This is where AI comes to the rescue. Qualcomm engineers have shown that it is possible to train a neural network to correct the environment’s interference with the 5G signal.
The training consists of placing QR codes that mark 100% accurate locations and using this “ground truth” to train the AI. Engineers then use the approximate 5G positioning as inputs to the neural network and compare it to the ground truth data to return incorrect results until the AI can no longer improve the accuracy of the 5G positioning.
It works out! There are cases where the Positioning accuracy dropped from 8 yards to 1 footwhich is a remarkable result.
Such accuracy is fantastic for consumer-level indoor positioning, but also suitable for industrial robotics. As more businesses look to deploy private 5G networks in their facilities, this is a great “bonus” feature that comes with 5G infrastructure.
While this is a very compelling proof of concept, we don’t know when you and I might need such a feature in our daily lives. The good news is that our 5G phones already have the necessary hardware. As 5G deployments become denser, opportunities for indoor location services will increase. We’ve never been this close to having a no-fuss and accurate indoor tracking system. For nerdy details, check out the video below.
Filed in . Read more about 5G, GPS, location and Qualcomm.
This article was previously published on Source link
