Class Reunion - Class of 1969 - Liberty, Mississippi
Lei Feng net (official account: Lei Feng net), it is not difficult to do automatic driving, but adding a word "full" in front of it is not a concept. However, when we are busy discussing technical details, the use scenario of autopilot is rarely mentioned. So who is the real audience of self driving cars? How can we satisfy the practicability and economy and get strong user stickiness at the same time?
Obviously, the media, analysts and the industry should not ignore this basic problem when predicting the market trend of autonomous driving. Egil Juliussen, director of
IHS Automotive research, divides the use scene of autopilot into four categories: last mile distribution, automatic driving truck, fixed route automatic driving and Robo-taxi. Understand theNuances matter, after all, because they affect the performance, cost, and timeline of future landing vehicles.
Next, let Juliussen take us through different use scenarios to see who is the head player in the four major types, the current state of the market and the progress of your testing and R & D work.
Also contributing to this article is Erez Dagan, senior vice president of product and strategy at Mobileye. Earlier last month, the Israeli giants ate moovit. What does electric cargo truck Mobileye really like about it? In an interview with
, Dagan stressed that Mobileye was aiming to "automatically drive cars into public transport."
Dagan is right. Autopilot can't run in a vacuum. Its value must be estimated by the travel and autumn condition of the whole society. Companies like movevit that master users' travel habits are, of course, the key to mobile Ye's future integration of real-time information and data.
Among the four types defined by Juliussen, "the most difficult one is Robo taxis".
Ironically, the head of autonomous technology giants, including waymo, cruise, Argo and Aurora, have a special interest in the Robo taxis market. At present, waymo one has landed in Arizona, while anbofo chooses gambling city and GM San Francisco.
AutomaticDriving test has become more popular. California alone has small electric truck gathered 65 autonomous driving companies with test licenses, while Argo's teams are based in Pittsburgh, Miami, Detroit, Palo Alto, Washington, D.C., and Austin.
Juliussen also pointed out that in the Robo taxis field, the United States has the leading position in wind investment, automatic driving development and testing, while China is in hot pursuit. "As for the EU, it's completely behind Robo taxis in autopilot testing."
In view of the strong gold attraction ability of Robo taxis, the market also recognizes its huge potential.
However, the future of Robo taxis is bound to be plagued by a huge social problem, that is, what if Robo taxis actually kills people at work? Who are the victims' families suing? What is the amount of compensation? Will the jury sympathize with a taxi without a driver?
A: if there are no more drivers on Robo taxis, the OEMs of any casualty accident will have to swallow the bitter fruit alone, because they can't throw the pot. Clearly, Robo taxis will change all existing areas of responsibility.
In the past, automakers have been slow to accept new safety measures in public demand, but what about autonomous driving? They must always pay attention to safety issues, because any fault manufacturers have to bear all the responsibilities.
According to Juliussen,Among the four types, the most easy to implement is the self driving truck, which is also accepted by the truck industry. In a recent interview, Plus.ai Coo Shawn Kerrigan stressed that trucks will be the first growing segment of the autonomous driving market.
Self driving trucks are specifically classified because of their unique business model. "The cost will be on our side." Kerrigan explained. Self driving trucks have the ability to save truck operators a lot of money because they no longer need drivers (the lack of truck drivers is also a common problem in countries around the world). However, it's only a start to get rid of drivers from the equation, and there are huge advantages in operating costs for self driving trucks. First, they can work 24 hours a day; second, self driving trucks are safer and accident rates are significantly reduced.
Juliussen points out that technically, the "freight center freight center" model is also easier to manage because "self driving trucks spend 90% of their time at high speed.". Given their low demand for urban driving, there are fewer "social issues" (i.e. interactions with pedestrians and other vehicles) to deal with.
Juliussen and Kerrigan are at odds on the size of the market for self driving trucks. The former thinks "this is just a small market", but the latter thinks its market prospect is broad. After all, the annual revenue of American truck industry is more than 600 billion dollars.
AutopilotThe success of a truck start-up depends on its business model. For example, the development of full stack automatic driving software Plus.ai They don't build cars, and they won't have a team that will run them in the future. Their mission is to provide key structural units for autonomous trucks.
Last fall, Plus.ai In this business model, Dongfeng Motor will successfully join hands with Dongfeng Motor, and the new joint venture will develop self driving trucks for the Chinese market. It should be noted that the first product of the joint venture company is based on Plus.ai The L4 stack can only realize L2 semi-automatic driving after it is put into use.
In other words, although everyone is interested in self driving trucks, it is not easy to think of L4 self driving trucks. Kerrigan points out that fully automated trucks are at least four years away. "We have to make sure the vehicle can handle all the extremes."
If the reality is so grim, how many years can self driving truck startups last without a real product?
Starsky robotics answers this question with bankruptcy. After five years of financing, they can't survive.
According to Juliussen, starsky has taken a non mainstream path. "I like them, though." He said. For example, starsky's trucks can be operated remotely at high speeds and in cities. However, "their ambition needs complete infrastructure to support, which may be worseInvestors ".
Kerrigan also said his company has raised $100 million. However, there are many competitors in the market. Juliussen's list includes embark, Ike and Kodiak.
At the same time, Tucson technology, which is supported by UPS, put forward a new plan in March this year. They will cooperate with ups to expand their own freight test project, and the modified L4 self driving truck will be incorporated into the fleet.
In Europe, Daimler acquired torc robotics last fall, and this year began to expand autopilot testing on US roads.
Of course, we can't forget the leading suppliers of autopilot stacks, such as waymo and aurora. In addition to Robo taxis, they will not let go of the cake of self driving trucks in the future.
fixed line autopilot
to say that most people have seen the "moving" automatic driving vehicle, I am afraid it must be a fixed line shuttle on campus, airport, hospital and theme park. Although
is deployed at the earliest time, it's the most boring four types of self driving cars, because they are slow and afraid of their feet.
Self driving shuttles usually look like ovens on wheels, with 10-15 seats crammed into them. Although they are more crowded than taxis, they can't compete with buses.
For those who like to wave and stop the taxi service, automatic drivingThe experience of driving a shuttle is no different from that of public transportation. But in Juliussen's view, in the future, they will play an important role in industry, such as the transportation of ports or mines.
investors are still interested in fixed line autopilot, but Juliussen thinks this category is the weakest investment because market demand will fluctuate greatly due to geographical location. In short, in a city where public transport is better, fixed line automatic driving is easier to eat and drive.
in the United States, such a country with weak public transport, the automatic driving shuttle vehicle is certainly not Robo-taxi. In countries with more public transport, people will have a higher degree of acceptance of fixed route self driving cars. However, since the public transportation is perfect, why another set of self driving shuttle?
Of course, self driving shuttles also have the potential and space to tap the market. Take the rural areas of Japan, for example, where there are no buses because of the continuous outflow of population, but the people who stay here still need to travel. In this case, the self driving shuttle can play a better role.
For the public transport system, profit is never its goal, and the subsidy from taxpayers is the real ammunition for its continuous operation. Therefore, in the real competition, I'm afraid the self driving shuttle is not its competitor. Erez Dagan of Mobileye believes that to survive, shuttles must "actively integrate into the public transport system.".
MobileyE chose to buy moovit because they believe that understanding people's travel habits can help shuttle operators inject more wisdom into their own services. Dagan asked, "what if operators can dispatch shuttles according to users' needs to make them appear where users need them most in time?"
In other words, the future direction is to coordinate the automatic driving shuttle with other travel options to meet people's travel needs from a macro perspective.
To talk about the leading companies in this field, the U.S. market looks at may mobility, local motors and Apple's acquisition Drive.ai 。 In Europe, there are two world-class top players, easymile and Navya.
Juliussen points out that of the four categories of self driving, the last kilometer of goods delivery is the fastest growing and most well financed one.
In this category, nuro is the most famous. Last year, the company made nearly $1 billion in investment from Softbank, and in retail, it pulled in the giants Kroger and fry's food.
In addition, nuro has licensed its own technology to third-party companies, such as autonomous truck company Ike.
Earlier this year, nuro was authorized by NHTSA to deploy 5000 R2 delivery electric vehicles on the road to deliver food, groceries or other services to users.
JUliussen describes R2 as "only half the size of a car, but the height is still prominent in the traffic flow.". It's clear that R2 doesn't have any space for the driver. "
Not all U.S. states allow such delivery robots to go on the road, and even if they can, they are required not to exceed 6 miles per hour (about 10 kilometers per hour). In addition, this kind of car can only run on the sidewalk, and when encountering bicycles and pedestrians, they should consciously yield. At the same time, the operating company also provides remote operators for the delivery robot.
Recently, giants have begun to infiltrate this field. Amazon has developed a sidewalk autopilot, and WAL-MART has launched a grocery delivery service with Udelv. In China, neolix, the strategic partner of Baidu Apollo platform, has received US $28 million in a + round financing.
Of course, the most attractive thing for investors is the cost. After all, they need much less manpower, and the vehicles are much cheaper than the delivery vehicles.
ei Feng net
the original articles of Lei Feng net are not allowed to be reproduced without authorization. Please refer to the reprint notice for details.