Self-driving cars are vehicles that operate automatically and need minimum human intervention to control the functioning. They use highly advanced sensors like sonar, radar, GPS along with the backing of advanced software to interpret sensory data in order to locate paths and obstacles.

History of Self-driving cars
Though the experiments for self-driving cars are being conducted since 1920, the first automatic car was launched in the year 1977 by a Japanese mech laboratory Tsukuba. It tracked the white lines marked on the street and had 2 cameras which used an analog prime computer to process signals. The first prototype for self-driving cars was launched around 1980 by Carnegie Mellon, followed by DARPA in 1984, and Mercedes in 1987. But the DARPA Urban 2005 and 2007 challenge changed these self-driving cars into an absolutely possible reality. By then, the cars could do a lot of things automatically like taking U-turns, parking within a slot, following traffic rules, and even maintaining lanes. In 2009, the biggest project for self-driving cars was launched by Google, followed by Elon Musk from Tesla. Even cab services like LYFT and UBER have made efforts in this direction.

How do they work?
Though still in its primary stage, the technology of cars is getting popular and has the potential to transform the transport system completely. Many companies like Google, Tesla, Uber, Nissan, etc, have invested millions in technology.

The various levels of autonomy:
There are various levels of self-driving cars (starting from Level 0 to 5)

• Level zero
  As the level suggests, these are the cars with zero level of automation. The driver is responsible for car control.
• Level one
  These vehicles have more than one system automated, like braking, steering, or acceleration. But the system does not work in coordination with each other. Some examples could be automatic brakes in case of emergency, cruise control, etc.
• Level two
  These level two cars can control some basic functions simultaneously like steering and speed control without human intervention for a short duration. Eg -Tesla.
• Level three
  These are almost fully automatic cars under all major safety conditions. The system alerts the driver to take control if it fails.
• Level four
  Fully automatic cars that require no human intervention. If the system fails, the car stops automatically following all safety standards.
• Level five
  These are the cars of the future. They would be designed top to bottom to function completely without any intervention at all. They would not need drivers, even with the controls in place.

All self-driving cars have a map which helps them identify their surroundings. UBER uses laser sensors to create this map, while Google is said to use a combination of sonar, radar, lasers, and cameras with high power. This input information is then processed by the software to generate instructional signals to the car’s actuators, which enable the control of steering, braking, and acceleration. Various predictive algorithms, code-based rule engines, object deciphering algorithms differentiate between two objects based on their shape, size, and volume like a stone and a cycle, and facilitate the software to navigate through the traffic. Semi-automatic cars need human intervention if the system falters.

Another distinguishing factor for self-driving cars is the IOT (Internet of things). Some models are equipped with technology which enables these cars to talk to other vehicles and even communicate with satellite systems to decipher the destination or its current location. These cars are smart and reprogrammable. The continuous enhancements to vehicle software due to reprogramming make it smarter with time.