blog posts

The car is a mythical concept and its realization in the future! In 2015, the Guardian newspaper predicted that by 2020, 10 million cars would be on the streets! The forecast was based on General Motors, Google, Toyota, and Honda reports. But now, the year 2020 has arrived, and there is no news of self-driving cars on the streets! Despite many efforts by the world’s innovative companies such as Tesla, the car is still out of reach, except in some experimental cases. Of course, some self-driving cars do things like automatic braking or keeping the car between lanes; Some cars, such as the Tesla Model S, even allow automatic driving on the freeway; But we are still far from fully self-driving cars.

What is a car, and how does it work?

The idea behind cars is very simple: Engineers install multiple cameras around the car that monitor everything around it and block it if it tries to hit an obstacle. We teach the car’s internal computer what the road rules are and leave the driving to it! This simple explanation has many complexities when implemented. Driving is one of the hardest things a person can do routinely. We humans do not just drive by knowing the rules of the road. We use many tools; For example, we use eye and gaze movements to understand the right of way or apply the weather conditions in our driving. We sometimes make calculations that do not easily fit into the rules and regulations.

Google Car (waymo) 

Simple tasks while driving, such as identifying objects, are a challenge for the car and its computer. Waymo (Google’s sister!) Is one of the leaders in the automotive market. The company uses high-resolution cameras and the LEAD method to measure the car’s distance from the objects around it. This method measures the distance by shining a laser light on the target object and analyzing the reflected light.

A self-driving car uses multiple cameras and sensors to create a complete picture of everything around it, including cars, bicycles, pedestrians, obstacles, and more. This map also shows how each object moves. To predict how objects move around a car, we need large amounts of data to teach motion prediction, artificial intelligence models. This data is obtained after traveling millions of miles by car. Gathering all the data needed in real life is hard work. For this reason, engineers obtain large volumes of data through simulation; But the challenge is to make sure the simulated data is the same as the real data.

What is the reason for the time-consuming construction of self-driving cars?

The performance of the car is based on artificial intelligence. The decade 2010-2020 was a unique year for artificial intelligence! Many advances were made in translation, sound making, computer vision, and object recognition during this decade. At one time, it was not easy for artificial intelligence to recognize the image of a dog; But now it is not a difficult task at all. In the middle of the last decade, the huge advances in artificial intelligence have made individuals and companies very optimistic about the future of self-driving cars. Researchers have introduced very new horizons to the world community. In practice, when building a self-driving car, the limitations of artificial intelligence became clear to us. Despite the great time, money, and effort that has been spent so far, none of the world’s leading companies have yet achieved completely reliable products.

Many of the problems we face in developing self-driving cars require large amounts of data. The best way to teach a car is to show it billions of hours of driving video; In this way, a computer-based on artificial car intelligence can become a good driver! Modern machine learning systems work well when they have a lot of data; But if they do not have a lot of data. Data collection is expensive for smart cars. On the other hand, because some conditions, such as an accident, are special conditions, the car may not be ready for it; Because he did not receive enough data about it.

Various car manufacturers have made several attempts to solve the problems. They have been collecting a lot of driving data for hours; They have even simulated different conditions to make the car more experienced. Of course, there have been some successes; Waymo cars are now roaming the streets of Arizona, USA, without the need for a driver! They will probably move to more cities by the end of 2020.

What is our perception of a world full of cars?

Despite the very serious challenges ahead, many power companies are still struggling; they are not disappointed because the acquisition of self-driving cars will create a huge change in the world and bring a lot of wealth to its manufacturers. Imagine a world where you have to sit in your car, and the car does the rest for you! Or you can even take a taxi online; A taxi without a driver! The car itself also helps a lot to people with physical disabilities.

Some experts believe that with the advent of self-driving cars, car ownership will change, and many more people may not buy cars. According to a recent survey, about 9 percent of Americans say they buy one immediately after buying a self-driving car; About 38% said they would buy the cars after a while; But more than half of the people do not want to buy a car! Of course, these percentages are expected to change quickly after the successful launch of the first cars. Another important point is that to have a real-world full of self-driving cars, governments must make the streets more suitable for the passage of such cars.

Google and General Motors cars Tesla cars are not among the best.

Almost all of the world’s leading automakers have made a series of minimal efforts to achieve self-driving cars. But some of them are much more serious. To evaluate the success of a self-driving car program, we need to consider three main factors:

• The amount of investment for the project.
• The amount of distance that the car has traveled is a measure of the progress of its artificial intelligence model.
• Disengagement index: The number of moments that the car needs a human user in each mile of driving and can not control the situation itself.

In all of the above indicators, Waymo, a subsidiary of Alphabet (owned by Google), is ahead of its competitors. The company recently announced that it had driven about 20 million miles. The disengagement index announced for this company is 0.09 per 1000 miles. Next in line is General Motors’ cruise car. The car has driven about 0.5 million miles, and its disengagement index is 0.19 per 1,000 miles. According to experts, Waimo and General Motors are the leaders in the automotive market.

Car killer cars!

March 18, 2018, was the first time a car overtook a pedestrian! An Uber car in which the driver was also killed a 49-year-old American woman riding a bicycle in Arizona. This proved to everyone that the self-driving car still has a long way to go. Of course, some people say that just as a person makes a mistake while driving, a car can also make a mistake; It should be noted, however, that for every 100 million miles driven by humans, there is only one fatal event. Waimea has traveled 20 million miles so far; Of course, it was not accidental; But it is too early to judge the safety of smart cars. Uber has not yet commented on the distance traveled by its smart car, But it is probably still more of a human error than a human error.

Tesla car

Tesla is the largest manufacturer of electric cars in the world. The company was founded in 2003 by Ilan Musk and several others. With the acquisition of startup SolarCity, Tesla was able to take a decisive step towards using solar energy to charge its cars. Tesla currently has four main products:

• Model S
• Model 3
• And Model X.
• Model Y

In 2014, Tesla began offering self-driving cars. In 2016, the company made improvements to the sensors used in its vehicles, and since then, all Tesla vehicles have been equipped with a pseudo-car system. Tesla’s Autopilot hardware includes:

• Eight cameras
• 12 ultrasonic sensors
• Forward-facing radar

Tesla car intelligent car systems include:

• Adaptive cruise control (to adopt a safe speed given the distance from the car in front)
• Offline warning
• Emergency braking
• Semi-automatic steering
• Double Park and L Park
• Call the car from the parking lot

According to the explanations provided, the car industry still has a relatively long way to go.