Tesla just unveiled its new autopilot system. It is a lot more than just assisting drivers to drive safely. Machine learning technology represents the cutting edge algorithms whereby computers look at large sets of data to analyze it and make decisions based on it. Many big software companies like Google and Facebook have been working on these technologies for years to eliminate the need for human assistance and make things faster.
Machine learning enables computers to develop artificial intelligence. Although, Elon Musk in the past has been wary about uses of artificial intelligence for things that are potentially harmful to humans and co-wrote a letter with more than 1,000 AI researchers to ban killer robots. He is all up for utilizing AI for beneficial purposes.
There are many other automakers working on AI for vehicles. Here is how Elon Musk differentiated Tesla’s AI from what other companies have to offer.
“The whole Tesla fleet operates as a network. When one car learns something, they all learn it. That is beyond what other car companies are doing,” The system will learn from the drivers who are actually driving cars on autopilot. There is a lot of controversy surrounding the use of AI to build self-driving cars. However, Google is already working on this technology to make this happen. Tesla’s self-learning autopilot system can be compared to the one developed by Nest’s learning thermostat which relies upon sensors and algorithms to learn about its driving environs.
The work on enabling computers to take over important parts of driving starts by training algorithms by providing large sets of data to them. The companies can use from a few hundred to millions of videos to help build vocabulary for the computers to learn from. The idea is to cover all possible driving scenarios and the way to act in them to avoid a potentially dangerous situation. Companies work on collecting as much data as possible to make sure that computer’s working on managing these kinds of situations become smart and make intelligent decisions for any given situations. In order for the cars to work, they need to be fitted with plethora of sensors and cameras as well as a radar system which continuously feeds live data necessary to make the autopilot work. The way the companies work to achieve safe driving via autopilot may differ by relaying on different types of data and using varying approaches.
Tesla is working on getting more precise mapping information than the one that is available as standard. The mapping system used by Tesla provides 100 times more detail than the standard mapping systems used for autopilot. The level of detail involved would give Tesla’s maps an edge over what is available in the market and make autopilot even safer than the ones to be offered by other companies.