Note: I'm very behind in my first inquiry! i'll be catching up over the next week or two. This is still my first one!
Which companies are working on producing driverless cars?
AB Volvo - AB Volvo is a Sweedish manufacturing company that focuses primarily on cars. Known also as Volvo Cars, they have teamed up with Uber- a tech company famous for their community-driven transportation app of the same name. Their goal is to design and finance autonomous cars for the mass market together. Just a few weeks ago, on November 20th, 2017, Uber announced their plan to purchase approximately 24,000 cars (compatible with self-driving technology) between the years of 2019 and 2021.
Tesla - Tesla is an American company that focuses on technology such as solar power, batteries, and cars. Well-known in the tech industry for making battery-powered cars not only well-known, but relatively affordable. As leaders in the world of future tech, the company is currently offering fully-functional self-driving cars in two models of their cars. I'd like to look into how they are programmed soon.
Intel, Waymo, and Mobileye - The close-knit trio of autonomous car research and development, the group began in 2009, with the Google-created 'Waymo' company joining the effort to create autonomous driving tech. In mid 2017, Intel purchased Mobileye- an Israeli company already focused on autonomous driving technology- and has continued to develop with it. That being said, Waymo has actually stated that they've been working with Intel since 2009- the very beginning of their company- meaning Intel has actually been in the race for some time just not directly. This was only revealed on September 18th, 2017.
There are other companies working on the technology and design of autonomous vehicles, however those three interest me the most. i'd like to look into Google's work later, though!
What kinds of programs make autonomous cars work?
While finding specific companies' documentation on which components and programs go into their driverless vehicles, there are a few general parts used in almost all of them.
For starters, the algorythem SLAM- meaning 'simultanious localization and mapping'- is used to find obsticles around the vehicle both via sensors/cameras, and through an offline map (that routinely updates with environmental changes). You can think of it like the eyes of the car, with a built-in map. The biggest difference is how the vehicle can access both silultaniously, and with more accuracy and speed than humans can.
DATMO, meaning 'detection and tracking of other moving objects', is another algorhythem being fused together with SLAM to add other cars, pedestrians, and other moving objects into the mixing pot of what the car can track (and avoid). It is being developed by Google.
Neural networks, which enables a computer to engage in 'deep learning', are also used in autonomous vehicles. These networks are used to allow technology to 'learn' as it operates. They imitate the way the human brain builds links between bits of information to create associations between them, and 'gain knowledge' through repeated observed occurences. Layers of stored info make up the 'brain' of these neural network-infused computers.
GPS systems are often used to allow passengers in driverless cars to input the location they'd like to travel to, so that the car can decide on the fastest route- taking traffic, weather, and closed roads into consideration.