In this blog post, we will examine whether driverless cars can be safer than human drivers by looking at the driving principles and security technologies of autonomous vehicles.
The field of driverless cars is actively being researched alongside the development of artificial intelligence. In particular, autonomous driving technology is attracting attention for its potential to reduce traffic accidents and congestion and provide efficient transportation. However, the concept of driverless cars has not yet been commercialized. In addition, the psychological factor of not being able to control the car oneself increases anxiety among consumers who use driverless cars.
Anxiety about driverless cars also stems from real-life accidents. In 2016, an accident involving a driverless car based on Google’s artificial intelligence and a fatal accident involving a driver in a Tesla self-driving car have raised concerns among consumers.
In addition, on March 24, 2017, an Uber self-driving car was involved in an accident in Arizona during a test drive, colliding with a vehicle next to it and overturning, resulting in the suspension of operations. These accidents, which have occurred at several companies, have spread doubts about self-driving cars. However, these doubts about self-driving cars are based on misunderstandings and psychological anxiety on the part of consumers, without an understanding of how they work.
To correct this, it is necessary to understand the principles of autonomous driving and have an accurate understanding of driverless cars.
In order for a driverless car to drive, it must first recognize its surroundings using cameras, radar, GPS, and other devices. After collecting various data, all of the data is digitized and transmitted, and based on this data, the car decides which direction to go and at what speed.
Furthermore, in order to discuss the safety of driverless cars, it is necessary to understand the detailed driving principles. Driverless autonomous vehicles are divided into systems for recognizing the vehicle’s surroundings, systems for determining the vehicle’s absolute position, and systems for avoiding obstacles. They are equipped with cameras, real-time control computers, and computers that process image and laser data, and the computers that process the information collect information from the external environment and transmit it via a network. Like a child learning on its own, driverless cars repeatedly go through the processes of recognizing, understanding, avoiding, and processing, and through a functional driving assistance system, they perform driving, braking, and responding to unexpected situations.
Driverless cars are equipped with independently configured hardware sensors and software to maintain stability through the system, so that even if a camera or laser sensor is damaged and does not function, the entire system will continue to operate without any problems. In addition, a vehicle control system is equipped with a switch that stops the vehicle in an emergency to prepare for unexpected situations. Each piece of data is transmitted to a computer via a network, and since this data is extracted from the network according to each algorithm module, the algorithm module and sensor module are configured independently so that the system can operate normally even if some sensors fail.
Other factors contributing to the instability of driverless cars include hacking and cyber attacks. This is because the algorithms and data of driverless cars are influenced by artificial intelligence and computer programs. Recently, RSA encryption, which uses prime factorization, has become the mainstream method of encryption for cyber security, such as networks. The idea behind RSA encryption is to express important information as two prime numbers, then transmit the product of the two prime numbers with a hint to use as a password.
RSA encryption is the most common method. Recently, research is underway on algorithms that reduce the time required for prime factorization through various approaches. However, even with supercomputers, this takes a long time, so RSA encryption can delay attacks effectively. In response to this, quantum cryptography is emerging as a new encryption paradigm to create more secure and robust encryption.
Quantum cryptography is an encryption method that applies Heisenberg’s uncertainty principle, whereby when a quantum is observed from outside, its state, which was simultaneously 0 and 1, is determined as either 0 or 1. By tracking changes in the quantum, it is possible to determine whether someone is eavesdropping on the network, thereby protecting the Internet.
Autonomous vehicles still have some issues in terms of safety and technology that make them unreliable. In particular, technical factors are exaggerated by the media, which tends to focus on a few mechanical failures. However, the technology behind driverless cars is equipped with advanced safety features that enable them to drive more safely than ordinary drivers. They also have sophisticated security systems to protect against hacking and cyberattacks. In the future, it will be rare to hear about accidents caused by driverless cars.
In order for driverless cars to become commercially available, not only technological advances but also legal and ethical issues must be resolved. Legal mechanisms must be put in place to clearly define responsibility in the event of an accident involving an autonomous vehicle, and clear standards must be established for autonomous driving algorithms to make ethical judgments. Only when these issues are resolved will driverless cars be able to become an integral part of our lives.
