news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

In the baggage handling hall of Terminal 1 of Hong Kong International Airport, an unmanned trailer is moving passenger luggage. At the same time, in the restricted area of ​​the airport, an unmanned bus is picking up employees. Since Hong Kong International Airport and UISEE Technology (Beijing) Co., Ltd. (hereinafter referred to as "UISEE Technology") started unmanned operations in 2019, the unmanned "Blue Car" has gradually become a multi-tasking "AI driver" in the world's busiest cargo airport.

From Zhongguancun Science Park to Hong Kong International Airport, from exploring car manufacturing to empowering logistics, travel, smart cities, etc., UISEE has spent 8 years to carve out its own path in the autonomous driving track. A reporter from Beijing Business Daily recently visited UISEE to explore how it defines itself in the ever-changing market and how it can truly implement the "all-scenario, truly unmanned, all-weather" experimental scenarios in thousands of industries.

Don't build cars, build systems

In 2014, Tesla's Autopilot system was officially launched; in 2015, Uber poached more than 40 artificial intelligence and robotics experts from Carnegie Mellon University to develop autonomous driving technology... As the wave of artificial intelligence and autonomous driving became more and more fierce, a large number of domestic entrepreneurs began to flock to the field of autonomous driving in 2016. Wu Gansha, co-founder, chairman and CEO of UISEE, also founded the company in the same year.

In fact, before founding UISEE, Wu Gansha had been deeply involved in the field of big data for many years and was the chief engineer and director of Intel's China Research Institute. For Wu Gansha, who has witnessed the two waves of the Internet and smartphones, choosing to start a business is not about ideas, but about timing.

Wu Gansha, who seized the opportunity of autonomous driving, used his status as a celebrity founder to attract a number of well-known investment institutions to invest in the company, including China Science and Technology Innovation Star, Zhen Fund, and Sinovation Ventures. Moreover, within less than a year after the company was founded, the autonomous driving concept car "Urban Mobile Box" won the international award Red Dot Design Award. Subsequently, in 2019, through the unmanned logistics vehicle cooperation project with Hong Kong International Airport, the unmanned driving solution and service system were commercialized on a large scale.

In the early days of entrepreneurship, the company was in the limelight and full of fighting spirit, and the development of UISEE Technology was smooth sailing. However, if Wu Gansha and his team members want to still have a place after the tide recedes, they must remain sober and prudent and not take it lightly.

At the beginning, UISEE also thought about further industrializing the application of the autonomous driving concept car, but in the end it returned to its most advantageous part, providing the car with autonomous driving capabilities. Wu Gansha once said that UISEE has been constantly exploring and summarizing in practice, and ultimately it will obtain a large amount of low-cost data and real vehicle verification capabilities through passenger cars, and realize the sinking of specific application scenarios and the provision of overall solutions through special/commercial vehicles to obtain high gross profits.

In an interview with a Beijing Business Daily reporter, Wu Gansha mentioned the concept of "calculating accounts" many times. Taking the unmanned micro-circulation bus in the early years as an example, Wu Gansha told reporters, "The cost of building and operating such a car is far greater than the cost of hiring a driver to drive and operate it, and it can only replace one driver. This account is hard to calculate. Later, the unmanned logistics vehicle project at Hong Kong International Airport mainly provides autonomous driving solutions, and can also replace the three types of work of loading, hauling and unloading at once. It is obviously a commercial path that can be quickly implemented."

"We don't build cars, we only provide autonomous driving services. If you imagine a driverless car as a computer, what UISEE does is the operating system." Wu Gansa said that the car is just the carrier for UISEE's operating system. At this stage, the company's goal is to make a safer autonomous driving system that can match more car models and scenarios.

From 99% to 100%

If the concept of "making systems instead of cars" has clarified the advantageous direction for UISEE, then the cooperation with Hong Kong International Airport is an important turning point for it to quickly open up the market.

Wu Gansha pointed out that it was not a "headstrong" decision to start the airport business at the beginning, but the entire airport environment has more stringent requirements for driverless cars. Once this level is passed, it will not be difficult to enter other scenarios. Just like in a war, you must first seize the high ground, and then it will be relatively easy to take over other territories. At the same time, for UISEE Technology, the Hong Kong International Airport is a rare angel customer that has both the pain point demand for logistics and transportation and is willing to innovate and try with start-ups.

According to data released by the Airports Council International, Hong Kong International Airport will handle 4.3 million tons of cargo in 2023, making it the world's busiest cargo airport in 2023. However, the "busiest" airport also bears huge pressure on logistics and transportation. The traditional airport logistics operation mode that relies on manual driving faces many challenges such as rising labor costs, labor shortages, and high safety risks in the operating environment.

In 2019, Wu Gansha received a request from Hong Kong International Airport to remove the safety officers on driverless logistics vehicles. Wu Gansha told the Beijing Business Daily reporter that this meant leaving his original comfort zone. There was no way out and he had to fight. "Earlier, many autonomous driving companies, including ours, equipped driverless vehicles with a safety officer for safety reasons so that humans could take over special situations that might be encountered. But this does not 'quench the thirst' for customers. Back to reducing costs and increasing efficiency, only by completely removing the safety officers and achieving full-scene, all-weather, and full-process driverless driving can the requirements be met. In this race, achieving 99% is equal to 0, and only achieving 100% is qualified."

How to go from 99% to 100%? Wu Gansha said that it is all about focusing on the last "1%", which is also the real difficulty in the commercialization of autonomous driving.

Take conquering all-weather as an example, that is, to break through 1% of extreme weather. "Every drop of rain recognized by the autonomous driving system is an obstacle. And the raindrops in different places are of different sizes, some are dense, and some are sparse. For example, Hong Kong is geographically close to the sea, and the rain has a high salt content, and is accompanied by salt fog and high temperature weather." Zhang Dan, director of the basic platform research and development department of UISEE Technology, told the Beijing Business Daily reporter that only with sufficient sample data and training can the system accurately judge the external environment.

To this end, UISEE Technology has also gone through a process of "waiting for wind and rain - creating wind and rain - calling the wind and rain". The company first went to various parts of the country to collect various data for testing, which is the stage of waiting for wind and rain. Later, because the cost of waiting was too high, it began to use sprinkler trucks to create wind and rain. With the introduction of more advanced technology, it began to simulate various types and intensities of rain and fog by editing rain and fog generators, and finally achieved "calling the wind and rain".

"At present, in rainy and foggy days, as long as people can drive normally, our driverless cars can be on the road, and can perform almost the same as having a real driver in the car." Zhang Dan gave an example, saying that when its unmanned logistics vehicles are operating in moderate rain, heavy rain and rainstorms, they will activate the adaptive dynamic adjustment mechanism, analyze the data of the two sensors, camera and lidar, in real time, and make decisions within milliseconds to adjust the vehicle's operating speed in time to adapt to the current weather conditions, reducing the speed to 80% of the original speed in moderate rain, 70% in heavy rain, and 50% in rainstorms.

It is reported that before being approved to operate without safety personnel, UISEE actually went through three batches of charges. The whole process was also internally called the "Dungeon Plan". The R&D team of dozens of people broke through 1% of extreme weather, 1% of complex scenarios, and 1% of operating processes with human participation through closed training at the experimental base. In December 2019, the "truly unmanned" unmanned logistics vehicle project was officially put into operation at Hong Kong International Airport. At the same time, with this project, UISEE took the lead in guiding the introduction of a number of industrial policies and industry standards, and further expanded the results to important domestic and foreign airports such as Guangzhou Baiyun Airport and Singapore Changi Airport with the help of the "Belt and Road Initiative".

Towards a universal large model

In Wu Gansha's view, true greatness is not designed but evolved. This is also true for UISEE Technology, which is running on the road of autonomous driving.

After the success of the "Safety Officer Removal" project with Hong Kong International Airport, Wu Gansha originally thought that he could quickly expand the scale, but in fact there are still many areas that need to be optimized and improved in the operation and maintenance process, so in the next four or five years, the company has been constantly launching new products and upgrading old versions. Just for scenarios such as airport driverless driving, UISEE Technology invests more than 100 million yuan each year.

After the airport, UISEE is still exploring the operation of unmanned logistics vehicles on a larger scale. Wu Gansha introduced that industrial logistics is the company's second home. Unlike airports, industrial logistics has to face more fragmented and diversified requirements. "Airport logistics may be solved with one solution and one vehicle model, but industrial logistics has many subdivided scenarios and different requirements. In order to adapt to different scenarios and requirements, we have to do more."

At present, UISEE Technology's L4 unmanned logistics tractor products mainly include unmanned tractors, unmanned delivery vehicles, and unmanned flatbed trucks, which are mainly used in the transportation of production materials and goods in factories and parks, as well as cargo and luggage traction at airports. They have the ability to autonomously plan driving routes throughout the entire process to achieve autonomous obstacle avoidance, automatic parking, automatic unhooking and other functions, and can also be accurately connected to third-party equipment such as forklifts, platforms, rolling doors, etc., without the need for manual operation.

It is reported that UISEE has laid out four major sectors: passenger cars, unmanned buses, unmanned logistics, and smart city services, and has implemented autonomous driving technology in the fields of automobile manufacturing, hazardous chemicals, food processing, agricultural breeding, heavy industry manufacturing, civil aviation airports, industrial parks, etc. As of March 2024, UISEE has implemented more than 40 smart city travel projects, covering more than 100 factory logistics and distribution projects, and expanded more than 10 civil aviation airport transportation projects.

From airport heights to industrial logistics, from safety officers to "all-scenario, truly unmanned, all-weather" unmanned driving, in the process of UISEE's continuous development, Wu Gansha also found that designing unmanned driving solutions for a single scenario has an upper limit and is too costly. The most important thing is to be able to provide an autonomous driving operating system for all scenarios, that is, a universal autonomous driving model. "The road is getting wider and wider. As a service provider that provides autonomous driving solutions, the ultimate goal must be a universal autonomous driving model. Only in this way can we better cope with complex and changing industries and markets."

"In the past eight years, we have been constantly understanding cars and application scenarios to consolidate and enrich the operating system. An autonomous driving company is a 'labor dispatch' company, and the 'AI drivers' sent out must be able to cover various driver's licenses such as A, B, and C." Wu Gansha told the Beijing Business Daily reporter that in the future, UISEE will introduce large models to further enhance the capabilities of the operating system, and build an ecosystem based on the general large model to implement more scenarios and models. In addition, it will promote key businesses in airports and industrial fields to go all over China and the world, not only in Hong Kong, China, but also in Singapore, the Middle East, Japan, South Korea, Europe and the United States.

Beijing Business Daily reporter Jin Chaoli Cheng Liang