Smart cars
Transferred from: New Smart Drive
How congested are the roads in the United States? According to Inrix 2019 data, five of the ten most congested cities in the world are in the United States. In particular, Los Angeles can be described as the peak of the "plugging" session, and has been "ranked" as the first city in the world for six consecutive years.
Figure 1: California peak road segments
Perhaps annoyed by the blockage, Elon Musk recently lamented on Twitter: the traffic jam problem is simply the "ultimate boss battle", even the most powerful human beings in the world can not overcome the heavy traffic.
Figure 2: Musk social media
In fact, this is not the first time musk has "feuded" with road conditions. As early as 2016, he "spat out" during a traffic jam, saying that "the traffic jam is driving me crazy, I want to build a tunnel, and now I have to start digging!" So there was one more tunneling company in the world, the Boring Company.
Figure 3: Musk's vision of the tunnel
This time, Musk gave a new solution, "Tesla should probably make an artificial intelligence vision device that can connect traditional traffic lights." He further added that the device can only view traffic conditions and automatically maximize throughput. In the end, Musk summed up with some satisfaction that "self-driving cars can alleviate the traffic jam problem to some extent, and you don't have to experience this pain yourself."
Figure 4: Musk Social Media
As soon as the tweet came out, the onlookers speculated: the leader of bicycle intelligence - Tesla is also about to open a new era of vehicle-road collaboration?
In fact, if bicycle intelligence wants to achieve large-scale landing of automatic driving above L4 level, it may face the problem of excessive unit price of car manufacturing. After all, every car needs to be equipped with high-precision sensors, and the cost of the level will cause a certain limit on the market penetration rate of L4 vehicles.
Moreover, the L4 of the bicycle intelligent route also faces challenges such as security and ODD restrictions. Therefore, vehicle-road collaboration may be an important turning point for the commercialization of autonomous driving.
If Tesla can combine the intelligence of the car end + the intelligence of the road end with each other, it is another way out of high-level automatic driving. However, whether we can finally see Tesla's "AI traffic lights" is still a question mark.
Autonomous driving, pros and cons go hand in hand
Musk has long believed that self-driving cars could make traffic worse. "Self-driving cars will amplify traffic to the point of near-madness because you don't feel the pain of driving yourself."
The logic is very smooth, the emergence of self-driving cars has made travel easier, even people who can't drive can achieve "free travel", which will lead to a significant increase in traffic flow.
While autonomous driving will increase traffic, the advent of self-driving cars will reduce the phenomenon of "ghost traffic jams," one of the road conditions that drivers most dislike.
"Ghost traffic jams" are a wonderful traffic phenomenon. On a congested highway, it is likely that just because a driver brakes suddenly, suddenly changes lanes, or overtakes, causing a short pause, it will cause a series of pauses in the rear of the car, like a ghost. Even if the first car only takes 2 seconds to start after stopping, the time required for the final car to start is tens of minutes.
Studies have shown that if you are on a busy highway, the sharp braking of a novice driver can trigger a "traffic tsunami" that can affect up to 80 kilometers of road. In this regard, some Tesla fans said that it is time to let FSD stand out.
For his own AI, Musk has never hesitated to praise, "the world's best AI software engineers have realized how advanced Tesla's artificial intelligence is." In January, Musk claimed that the vehicles involved in the project had not been in any car accidents after Tesla launched the fully autonomous driving (FSD) beta version of the software a year ago.
FSD Beta is designed to follow the rules of the road. When self-driving cars become mainstream, they may be more "disciplined" than humans, more skilled in driving, and the situation of ghost traffic jams will be greatly reduced.
Musk has previously announced that Beta 10.12 is expected to be released in Canada on March 11. On February 27, Tesla has officially launched the FSD (Full self-driving) beta version in the Canadian market, which is the first market besides the United States to allow Tesla FSD to land.
Youtuber blogger WholeMarsBlog released Tesla FSD Beta 10 a month ago. Test video for 8.1. In the video, Tesla Motors, after turning on the FSD, is like a well-skilled and polite old driver, shuttling freely on the narrow city road.
But in the smooth autonomous driving journey of the vehicle, there are also weaknesses that cannot be ignored. For example, in previous FSD Beta tests, the vehicle turned right at 15 mph through an intersection and hit the protective column of the bike lane without warning.
Figure 5: The vehicle hits a baffle while the FSD beta is on
Coincidentally, a Tesla in the autopilot state, after encountering a green light and turning right as usual, almost hit a pedestrian who was about to cross the road. Pedestrians stopped when they saw the vehicle, and although the Tesla also slowed down, it had already pressed most of the zebra crossing when it slowed down.
Note: The vehicle nearly collided with a pedestrian
Experts believe that these problems are not easy to solve, and patching up a problem may create new complexities.
The prospects are good and the road is tortuous. It can only be said that fully autonomous driving has a long way to go.
Reference Links:
1.https://www.teslarati.com/tesla-self-driving-cars-will-amplify-traffic-elon-musk/
2.https://auto.sina.com.cn/jishu/2022-01-18/detail-ikyamrmz5826484.shtml
3.https://auto.sina.com.cn/news/hy/2022-03-07/detail-imcwipih7047080.shtml
This article is authorized to be reproduced, and the article only represents the views of the original author and does not represent the position of hard technology.
Hard technology
See, read and connect hard technology here. Focus on 8 hard technology industries such as optoelectronic chips, artificial intelligence, aerospace, biotechnology, and new energy, and rely on the unique advantages of scientific and technological information data of Wisdom Bud to let readers see technology trends, understand hard technology industries, and connect the future of innovation. Produced by Wisdom Bud Innovation Research Center.