|
Abstract
|
Road accidents contribute to the greatest number of deaths in the world. Deaths and injuries due to road accidents result in financial losses as well as physical and mental suffering. Even though a good driver is attentive enough to take sudden decisions, at some point, there is a requirement for an automatic decision-making ability in the car. Cars that can take prompt actions based on the environment without the driver involved is called a smart car. The car-following models are methods used in smart cars for accident avoidance. This paper presents an in-depth survey of various car following models based on IoT sensors, weather & road conditions, V2V networks, machine learning algorithms. A comparative analysis of multiple research articles with its techniques, merits and research gap is presented. Finally, the inference of the literature survey is provided.
Słowa kluczowe: V2V, Car-following models, accident avoidance, sensors, road accident.
Around the world, 1.35 million people die on road accidents each year. Road accidents are approximated to be the eighth major cause of death worldwide. More people now die in road traffic collisions than from HIV/AIDS or cancer. Even a small mistake by a professional driver can result in dire consequences. We can't solely rely on the driver to be aware of the car surrounding all the time. There is a need to provide the car with an ability to make decisions instead of the driver whenever there is a probability of a mishap. Car automation is the ability to use IoT sensors [1], mechanics, electronics and Artificial Intelligence [2] to aid a vehicle driver. A vehicle utilizing such features may be labelled as intelligent or smart. Researchers and Manufacturers have added a variety of self-operating functions to vehicles. Talking about IoT sensors, the first thought that comes to our mind is that, it is an instrument that detects and responds to a physical input from the environment. The specific input could be any natural phenomena like light, heat, motion, moisture, pressure, etc. for which we have sensors such as Ultrasonic sensor, DHT sensor, touch sensor, LDR (Light Dependent Resistor) sensor, IR (Infrared) sensor, etc. The result is data that is transformed into a human-understandable format for collection and data analytics. Car-following models [3] mostly use sensors to calculate the distance between vehicles, crash sensors, temperature sensors, tire pressure sensors, etc. Weather and road conditions also impact driving. In case of snowfall, the roads become extremely slippery and if a fast-moving car applies sudden brakes it might skid out of the road and crash. On the other hand, if the visibility is less due to fog, the driver won't be able to perceive the cars or people in front and won't be able to apply brakes on time that can obviously result in an accident. In order to tackle this problem, the car should be able [...]
V2V, Car-following models, accident avoidance, sensors, road accident. pojazd autonomiczny, czujniki, układy unikania wypadku
Language: en |