Every mountain opens the road, meets the water to build a bridge: everything is for commercialization.
Author | Jelly Ping
Edit | Wen Liang
Jobs once said, "Consumers don't know what they need, until we come up with our products, they find out, this is what I want."
But in the view of Tan Limin, founder and CEO of Nishii Technology, technology should serve the needs, especially artificial intelligence, and its essence should be to liberate human productivity.
"Technology inclusion and business value can be taken into account, but as a commercial company, commercialization is actually the most critical." Tan Limin said to the new smart driver.
A few days ago, Nishii Technology announced that the intelligent power exchange driverless commercial vehicle Q-Truck to achieve mass production, at the same time, its driverless brand Qomolo will also use new energy batteries across the line to achieve rechargeable, replaceable, energy upgradeable.
This is the first time in the world that the mass production of driverless commercial vehicles for power exchange has been realized.
With the blessing of the substation, the refueling time of the unmanned heavy truck in the port scene is reduced from 2 hours to 5 minutes, and its coverage area is also extended from the closed scene to the 200 km scene.
The imagination space of products and companies has also been greatly opened up at once.
"In the future, there will be two potential forces to support us to move forward and go deeper: on the one hand, Xijing will continue to take root vertically in the container logistics industry, on the other hand, it will expand the coverage around new energy heavy trucks." Tan Limin revealed.
From charging to power swapping
In fact, when Nishii Technology released Q-Truck in 2018, in terms of power supply, Q-Truck adopted the main and auxiliary battery system design, that is, the main and auxiliary battery system can complete 100% fast charging in 2 hours, and the auxiliary battery can be replaced in 3 minutes through a simple power exchange process to achieve a cruising range of 30 kilometers.
However, as an electric collector card, Q-Truck will still face a series of problems such as short mileage, long charging time, and fast battery decay when it is actually running.
In addition, Zhang Xiang, an analyst in the automotive industry, also told the new intelligent driving that due to the large load and energy consumption, in the case of the same mileage, the number of battery packs required for electric collection cards is more than that of passenger cars, and the cost is more expensive than traditional fuel vehicles, while the amount of pulling goods of vehicles will be limited, so before 2020, the attention of pure electric heavy trucks in the market has not been high.
Therefore, the power exchange is considered to be another energy supplementation solution that can solve the two core pain points of electric heavy trucks with short mileage and long charging time.
Chen Quanshi, a professor at Tsinghua University, has publicly stated that taxis, buses, logistics vehicles, docks, mines, transport vehicles at construction sites, etc., can only use fast charging or rapid replacement battery pack replacement methods to supplement electrical energy.
On the one hand, under the policy of "double carbon strategy", power replacement, as a new energy vehicle energy supply route promoted by the state, is making a comeback in 2021, becoming one of the hottest topics in the new energy automobile industry. OEMs continue to introduce power replacement models, and the number of new energy replacement related enterprises has surged.
In October 2021, the Ministry of Industry and Information Technology issued the "Notice on Launching the Pilot Work on the Application of the New Energy Vehicle Power Exchange Mode", and finally announced the launch of the pilot application of the new energy vehicle power exchange mode, proposing to achieve the goal of promoting 100,000 + power exchange vehicles and 1,000 + power stations, and set up 3 pilot cities with heavy truck characteristics.
"On the one hand, at present, the domestic steel, electricity, coal, cement and other industries are assessing carbon emission indicators, for enterprises, changing heavy trucks is conducive to reducing carbon emission indicators, on the other hand, the national subsidy policy on replacing heavy trucks also has a clear amount, so the current car companies will have the enthusiasm to declare and replace heavy trucks." Zhang Xiang thinks.
Tan Limin told the new intelligent driving that the future development trend of the port will undoubtedly be intelligent, unmanned and new energy.
Therefore, in December 2021, Nishii Technology also completely abandoned the "charging" and "replacement" and developed the Q-Truck, a new energy driverless heavy truck.
Tan Limin said that with the help of these two trucks and intelligent power exchange mode, Nishii Technology will replace the original charging mode and complete the power update within five or six minutes to maximize the vehicle's capacity coverage.
At present, the power exchange methods of electric vehicles mainly include chassis power exchange, side power exchange and back type power exchange.
For safety reasons, the Q-Truck version of Nishii Technology adopts the mode of side power exchange, and no longer needs to be manually replaced.
This is the first time that driverless commercial vehicles have adopted the model of replacing batteries on the inside of the cabin, and before that, almost all commercial vehicles on the market exchanged power, most of which used back-up power exchanges.
According to the introduction, the advantages of side power exchange, one is that the battery has equipment cabin protection, waterproof or the ability to reduce other potential risks, and the other is that the position of the battery is fixed in the cabin, so the acceleration and deceleration of the vehicle during use can greatly reduce the impact on the battery and the charging interface.
Pragmatic
From the perspective of Xijing Technology itself, Tan Limin believes that once there is a substation and a transit yard, the operation scope of new energy heavy trucks can be gradually expanded from the original closed scene to the 200 km scene, and then gradually expanded to 400 or even 600 km scene.
The business of Xijing Technology is mainly divided into two sections: "Smart Bulk Logistics Plate" and "Smart Big City Transport", the former takes container logistics as the carrier and faces scenarios such as seaports, river ports, railway/highway dry ports, etc., while "Smart Big City Transport" mainly aims at the digital city operation governance of urban digital upgrading and transformation.
However, according to Tan Limin, at present, Xijing Technology is still more focused on the logistics scene with containers as the central carrier, "nothing more than that it will shift from the seaport to the railway port, and then turn to the highway port to do some horizontal extension."
Therefore, in addition to releasing a power-changing version of driverless Q-Truck, Nishii Technology has also launched a power-changing manned networked truck E-Truck, which is not only competing with "unmanned driving".
From charging to power exchange, from driverless to manned, in fact, Nishii Technology has always been running through this "on-demand" style.
At the beginning of its establishment in 2015, Nishii Technology was mainly developing brain-like chips, and later began to shift to the application side of artificial intelligence from the perspective of commercial landing, choosing to directly define its own products from the needs of the downstream market.
In fact, at the beginning of the business, the starting point of Tan Limin's team was to use artificial intelligence technology to help humans improve production efficiency while finding opportunities for AI to achieve potential business value.
In his view, the essence of AI lies in liberating productivity, and when thinking about how to liberate productivity itself, a scenario is very naturally deduced - the field of production.
After doing a lot of research on the traditional manufacturing industry, and combining the market demand and its own technical characteristics, Xijing Technology saw the entry point of port automation upgrade.
In 2016, based on self-developed algorithms and chips, Nishii Technology used computer vision technology to analyze video stream information, and landed a set of intelligent tally systems that help ports automate tally, so as to identify the container number, box type, loading and unloading of containers and other states through cameras.
With the continuous communication with port customers later, Nishii Technology found that in the container logistics scenario, the unmanned horizontal transportation system will become a crucial link.
So in November 2016, Nishii Technology set up an autonomous driving team.
The so-called horizontal transportation system refers to the operation process of containers in the port from the shore bridge - collection card - bridge yard - gate, of which the collection card part includes the collection card / straddle carrier / AGV, which is also divided into internal collection card and external collection card.
According to the "China Autonomous Driving Industry Research Report" jointly released by Hanergy Investment Group and 36Kr a few days ago, the mainland currently has nearly 100 ports, including more than 10 large ports, and about 20,000 sets of containers in the port for container transportation, but only 13 ports have landed automatic driving cards.
The report predicts that with the maturity of automatic driving technology, the introduction of autonomous driving trucks in the port increase, the starting sales of automatic driving trucks will grow steadily, it is predicted that by 2023, the domestic port terminal autonomous driving truck sales will exceed 5,000, the market size of about 5 billion yuan.
Port driverless collection card has a broad market growth space, while the traditional cross-transport vehicle and AGV also have a lot of hard injuries, so that Nishii Technology feels that AI can achieve a breakthrough here, and automatic driving has become a typical application of artificial intelligence in horizontal transportation.
In an industry report released by Nishii Technology, Nishii Technology said that traditional cross-transport vehicles have these major shortcomings:
The capacity of the container is low, and the container can only be stacked on three to four floors at most;
High requirements for the foundation, especially at present, Some backfill land area terminals are being built in Asia, and if the traditional straddle carrier is used, the terminal will need to carry out special foundation reinforcement for the entire operating plant area;
The cross-transport vehicle body is high, the technical requirements for driver operation and maintenance personnel are high, and the cost is expensive, and the initial investment is large.
As for the current mainstream AGV, the technology used is magnetic nail guidance or base station guidance, the cost of peripheral devices is very high, and the operation efficiency is relatively low, and the effect of its promotion and deployment will also be affected.
Therefore, after The release of the driverless brand Qomolo by Nishii Technology in 2018, it has also successively launched the full-time driverless collection card Q-Truck, the intelligent cross-transport vehicle launched in cooperation with Zhenhua Heavy Industry, the automatic driving system based on the IGV model platform, the cross-scenario unmanned heavy-duty mobile platform Qomolo One, and the manned minibus Q-Shuttle.
Tan Limin summed up the two beliefs that pushed Xijing Technology to continue to move forward in recent years:
Pure technology does not form productive forces, only technology integrated into production is the real productive force;
When serving specific scenarios, enterprises can not only provide autonomous vehicles, but also need to provide a self-driving car that can work, which makes Nishii Technology need to have an in-depth understanding of the entire production process of customers.
Therefore, it is not difficult to understand why the series of products of Nishii Technology will undergo a series of changes such as switching the power platform from diesel vehicles to electric vehicles, switching from charging to power exchange, and the production mode of vehicles from cooperating with OEMs to making cars to now making their own cars, from driverless Q-Truck to human-driven E-Truck.
"As a commercial company, commercialization is actually the most critical." Tan Limin said this to the new intelligent driving.
The number of orders in January has exceeded last year's total
Tan Limin also believes that artificial intelligence or automatic driving are very broad concepts, "relying on a single technical point can not open up the industry, but need to link and integrate multiple technical points to form more quality and value products, in order to achieve the commercial landing and commercial value of these technologies."
From the perspective of technical guidance thinking, this means that enterprises need to complete the transformation from the thinking of engineers of algorithms to the thinking of engineering systems, the biggest difference between the two, from the result, one is the demo (0-1) that is still doing the prototype, and the other is for the purpose of mass production (1-n).
This is very closely related to the development stage of a team and the scale of the product it needs to deliver to the user.
Sun Zuolei, partner of the unmanned driving business of Xijing Technology, said that the user volume of Xijing Technology and the delivery demand it faces now pushes Xijing Technology to go out of the idea of doing demo and go directly to product engineering.
"The engineer thinking of the algorithm and the thinking of the engineering system, the underlying technical accumulation required by them is similar, but the latter is the complete set of the former, and the former is a subset of the latter." Sun Zuolei believes that starting from the thinking of the engineering system, enterprises need to consider many engineering issues when mass production of products, such as supporting services and technologies, supply chain system management, and various trade-offs in the process of integrating products.
Therefore, in the face of the port scenario, Nishii Technology takes the software and hardware self-research route, in addition to the application-side collection card, its products also include high-precision map construction, vehicle scheduling system, multi-field front fusion perception engine, self-developed road test V2X perception calculation unit, WellSim unmanned driving and customer scenario simulation platform, unmanned software system Q-Pilot, automatic driving chassis AVCU, remote steering Q-Wire, etc.
As we all know, learning-based perception methods are data-driven. Sun Zuolei said that the difficulty of data engineering is much higher than the development of deep neural network models itself, and there are high requirements for the infrastructure construction that supports data intelligence.
"Because our driverless fleets are all in a mixed state, we have been working hard to create perception capabilities that can adapt to open roads, and we are continuing to invest in data platforms."
At present, Nishii Technology has initially built a data closed-loop system driven by MLOps.
Tan Limin believes that independent research and development and manufacturing of software and hardware will not only help Xijing Technology grasp the pace of development, but also better product consistency.
In addition, Xijing Technology also said that at present, they are the only company in the world that can provide the global artificial intelligence transformation of the port, in addition to unmanned driving and intelligent tally, the smart port system of Xijing Technology also has smart gates, intelligent yards, smart port machine transformation, intelligent safety supervision and other functions.
Judging from the results, the progress of Nishii science and technology is also good.
Up to now, the products and services independently developed by West Well Technology have been put into operation in more than 80 ports and logistics parks around the world, and the overseas market will likely surpass the domestic business this year and become its main market this year, and then the business in Sweden, Denmark, Malaysia, Mexico and other places will also be landed.
In 2022, Xijing Technology will also simultaneously open unmanned power exchange business in specific scenarios in four countries.
At the same time, the company's business model will also change due to the mass production of autonomous vehicles. At this stage, the business model of Xijing Technology is divided into three types: one is to directly sell vehicles; the other is that the vehicle is self-sustaining for the company, providing services to users, and charging according to the number of each operation; and the third is the charging mode of power exchange services.
"In January this year, the number of orders we currently have in hand has exceeded the sum of last year's orders." Tan Limin revealed to the new intelligent driving that if this trend is predicted, the order growth rate of Xijing Technology in 2022 will be 2-3 times that of 2021.
In addition, in terms of financing, Nishii Technology has also received a total of 11 capital increases so far. Just in early January, Nishii Technology also announced the completion of a new round of financing.
It is worth noting that one of the leading investors in this new round of financing is cccc capital's CCCC Blue (Xiamen) Industrial Fund (hereinafter referred to as "Blue Fund").
The Blue Fund is cccc's first market-oriented equity PE fund, and Xijing Technology has also cooperated with Zhenhua Heavy Industries, the holding company of CCCC Group, to launch intelligent cross-transport vehicles.
Blue Fund said that through this strategic investment by CCCC Capital, CCCC Will provide assistance to Xijing Technology in the fields of port design and construction, dredging and container crane manufacturing.
However, the port as one of the fastest scenarios for the landing of autonomous driving technology, there are many competitors, mainly driverless startups that are good at AI algorithms such as mainline technology, Si Nian intelligent driving, Feibu technology, Changxing intelligence, etc., as well as engineering equipment manufacturers such as SAIC Group, FAW Group and Sany Heavy Industry, etc., in recent years, they have also launched and landed many related autonomous vehicles and solutions.
But Tan Limin believes that the difference of Nishii Technology is that it can provide global solutions to users, rather than single-point automatic driving technology, "We are actually taking a relatively different path, and there is no benchmarking company in the industry at present." ”
summary
Tan Limin is not from ai technology, before founding Nishii Technology, he was mainly doing marketing work, but this also made his perception of the most cutting-edge technology of science and technology more acute, and his ideas were more pragmatic.
He used the words "studious, brave, and humble" to describe himself, believing that when managing a company, you must first have a humble learning mentality, and the reason is also very pragmatic -
In the field in which Nishii Technology is located, it is necessary to integrate multiple technical points in order to form a quality product that can be commercialized, "in this process, it is necessary to continuously learn horizontally to help them form cognition and make better judgments."
END
"Unmanned" driving is encountering a "employment" shortage
Revolution in autonomous driving "perception"? Inspiration from the human brain and frog eyes