Driverless cars could be lent a hand by smartphones
(Mirror Daily, United States) – With the upcoming revolution of driverless vehicles, new smartphone tech can help self-driving cars for the purpose of safer navigation. The problem with the futuristic technology that will be installed in numerous vehicles is clear: safety. The question is how many people would trust them.
There are a lot of worries surrounding self-driving cars, which is why they are being tested over and over again, placed in difficult situations and observed. However, making them safe is only the first step. The second is to actually convince prospective buyers that they are truly secure. That will be an effort, one that will be proven through clear and concise proof.
Researchers at the University of Cambridge created two complementary systems for driverless cars that will assure safety through a smartphone. It’s a device that mostly everyone has, and it could be the perfect accessory for a driverless car. And, both systems are more accessible at a much cheaper price.
There are three essential questions that have to be answered by engineering marvel: where is the car, what’s around the car, and what the car will do after the first two questions are answered. The two new systems will provide with more accurate responses, and, thus, improve safety.
The first system answers the first question. It’s yet unnamed, but it has perfected the technology of GPS down to a point. While the GPS can fail, go slow, or become unavailable, this particular one will be usable on a smartphone everywhere and anywhere. In fact, it promises to even be more accurate. It will pin down the location within just a few feet, and even show the angle the user is facing.
It’s claimed to be better than anything a satellite signal can provide.
The second system, called SegNet will provide an accurate answer to the second question: what’s around the car? Used through the smartphone camera, it will provide image representations, colored differently based on what the object is. This means that pedestrians will have a certain color, roads another, along with fences, trees, and multiple other factors. The researchers stated that it’s more efficient, and less expensive than LIDAR.
According to Alex Kendall, one of the researchers, this the first time that they’ve developed technology that uses deep learning. They essentially taught the system to ‘see’. This was done by carefully mapping out each pixel in around 5,000 images. The system learns as it’s being used.
While it’s unlikely to replace the already existing technology being used in self-driving car, both of these could be of great use as a back-up. They do say that you can never be too careful.
Image source: inmannews.wpengine.netdna-cdn.com
John Birks
Latest posts by John Birks (see all)
- Canadian Teen Killed by Toxic Shock Syndrome on School Trip - June 29, 2018
- Donut Fries Coming to Dunkin’ Donuts Nationwide - June 28, 2018
- Kohl’s Hiring Workers for Holiday Season amid Labor Shortage - June 28, 2018
