Power System In Self Driving Vehicles.

Self-driving car

Self-driving cars sometimes called self-driving cars or self-driving cars are vehicles that use a combination of sensors, cameras, radar, and artificial intelligence (AI) to navigate between destinations without a human operator. is. To qualify as a fully autonomous vehicle, a vehicle must be able to navigate unsuitable roads to a given destination without human intervention.

Companies developing and/or testing self-driving vehicles include Audi, BMW, Ford, Google, General Motors, Tesla, Volkswagen, and Volvo. Google's tests included self-driving cars such as the Toyota Prii and Audi TT, driving over 140,000 miles of his roads and highways in California.

How self-driving cars work

AI technology powers self-driving car systems. Self-driving car developers use machine learning and neural networks, as well as vast amounts of data from image recognition systems, to build systems that can drive autonomously.
Neural networks recognize patterns in data that feed into machine learning algorithms. 

This data includes images from the self-driving car's camera, from which the neural network learns to identify traffic lights, trees, curbs, pedestrians, road signs, and other parts of a particular driving environment.

For example, Google's self-driving car project called Waymo uses a combination of sensors lidar  light detection and ranging - radar-like technology , and cameras to collect all the data these systems generate. to identify and predict everything around the vehicle. what those objects do next. This happens in a fraction of a second. Maturity is important in these systems. The more the system drives, the more data it can integrate into its deep learning algorithms to make better driving decisions

     Fig.no: 1

                                 

Cars with self-driving features

Google's Waymo project is an example of a self-driving car, which means it's almost completely autonomous. There must be a human driver, but it just disables the system as needed. It's not strictly self-propelled, but it can be self-propelled under ideal conditions. high degree of autonomy
Many of the cars available to consumers today have self-driving capabilities, albeit at a low level of self-driving. Autonomous driving functions that will be available in many mass-produced vehicles from 2019.
includes the following:

• Hands-free steering centers the car without the driver holding the steering wheel. Drivers should continue to pay attention.

• Adaptive Cruise Control (ACC) automatically maintains a selectable distance  between you and the vehicle ahead until you come to a stop.

• Lane centering steering intervenes when the driver crosses a lane marking by automatically pushing the vehicle towards the opposite lane marking.

Levels of autonomy in self-driving cars

United States National Highway Traffic Safety Administration (NHTSA) envisages six levels of automation, starting with Level 0 where humans drive, and ranging from driver assistance technologies to fully self-driving cars doing. The five levels following Level 0 automation are:

• Level 1: Advanced Driver Assistance Systems (ADAS) assist human drivers in steering, braking, or accelerating, but not simultaneously. ADAS includes features such as rear-view cameras and vibrating seat alerts that warn drivers when they are about to drift out of their lane.
• Level 2: His ADAS that allows the driver to remain fully conscious behind the wheel and continue to take the role of the driver, while simultaneously steering and either braking or accelerating.
• Level 3: Under certain circumstances, an automated driving system (ADS) can take over all driving tasks such as parking. In such situations, the human driver must again be willing to take control of the vehicle and remain the primary driver of the vehicle.
• Level 4: ADS can take over all driving tasks and, under certain circumstances, monitor the driving environment.
• Under these circumstances, ADS is reliable enough that the human driver does not need to pay attention.
• Level 5: Her ADS in the vehicle acts as a virtual driver and takes over driving under all circumstances. Human occupants are passengers and are never expected to drive the vehicle.

               Fig.no: 2

Self-driving car safety and challenges

Self-driving cars must learn to recognize myriad objects in their path, from branches and debris to animals and people . Other challenges on the road can be tunnels that confuse the Global Positioning System (GPS), construction projects that cause lane changes, or complex decisions such as B. Where to stop so emergency vehicles can pass.

       Fig.no: 3

The system must immediately decide when to slow down, take evasive action, or continue normal acceleration. This is an ongoing challenge for developers, with reports of self-driving cars hesitating or turning unnecessarily when detecting objects on or near the road.

This issue was highlighted in March 2018 when he died in an Uber self-driving car accident. The company reported that the vehicle's software identified a pedestrian, but considered it a false alarm and did not avoid it to avoid a collision.The accident prompted Toyota to temporarily suspend testing of its self-driving cars on public roads. but testing continues elsewhere. Toyota Research Institute is building a test facility on a 60-acre site in Michigan to further develop self-driving car technology.

Accidents also raise liability issues. Legislators have yet to define who is responsible if a self-driving car is involved in an accident. There are also serious concerns that the software used to operate self-driving cars could be hacked, and auto companies are grappling with cybersecurity risks. 4,444 automakers are subject to Federal Motor Vehicle Safety Standards (FMVSS), and NHTSA reports that more work is needed for vehicles to meet these standards.

In China, automakers and regulators are pursuing different strategies to meet standards and make self-driving cars a ubiquitous reality.
A self-driving car (sometimes called a self-driving car or driverless car) is a vehicle that uses a combination of sensors, cameras, radar, and artificial intelligence (AI) to move between destinations without a human operator. To be fully self-driving, a car must be able to navigate unused roads to a given destination without human intervention.

Companies developing and/or testing self-driving cars include Audi, BMW, Ford, Google, General Motors, Tesla, Volkswagen and Volvo. Google's tests included self-driving cars such as the Toyota Prii and Audi TT, which drove over 140,000 miles of his roads and highways in California.
How self-driving cars work

Refrences : 

https://spectrum.ieee.org/exposing-the-power-vampires-in-self-driving-cars

https://www.wired.com/story/self-driving-cars-power-consumption-nvidia-chip/

https://www.udacity.com/blog/2021/03/how-self-driving-cars-work-sensor-systems.html

Submitted by ; ENTC (div -C) 

roll no 27 : Jay Nannaware

roll no 30 : Pratik Ohol

roll no 32:  Abhijeet Padwal

roll no 35:  Yogesh Pande

roll no 36:  Abhay Pandhare

Guided by : Prof .Dr. Medha Wyawahare.