Autonomous Car Technology And Its Shortcomings As Of Now
Companies like Uber and Google have been testing driverless and autonomous car technology for the past few years. Recently, there was an incident, where a driverless car killed a woman in broad daylight. The autonomous car had a backup driver, which was not enough to stop the accident. According to the driver, the victim walked out in front of them in a flash. In fact, the backup driver was alerted by the sound of the collision.
There are several reasons for this accident. Firstly, the autonomous driving system might have failed to register her as a human or it failed to detect her presence at all. This technology has to reduce accidents. But, such malfunctions make us question the feasibility and progression of autonomous driving.
Autonomous or Driverless Car
A driverless car is similar to a programmed device, which follows set of instruction written by the author. This means that a driverless vehicle should be able to operate fully by itself without any human interaction.
These cars use a technology called LIDAR (Light Detection and Ranging) that monitors around 60 meters of the vehicle in real time. Similarly, this creates a virtual 3D map of the surrounding to gauge the upcoming obstacles. There will be additional sensors fitted on the rear and front end of the vehicle to measure the distance from surrounding obstacles, which includes other vehicles.
The vehicle will use GPS system to calculate real-time positioning. When it comes to decision making, the car will have a set of instructions to follow. Similarly, the AI will be able to make other actions in an emergency situation. However, as a security feature, a user can always override the decisions made by the machine at any given time.
In theory, driverless cars will reduce accidents, as it will be programmed in a certain way. In a dangerous situation, these cars will negate the human error. However, the recent incident in Arizona means that the technology is not yet fool-proof for the real world.