Founded in 1987, Huawei is a leading global provider of information and communications technology (ICT) infrastructure and smart devices. Huawei has approximately 208,000 employees and operates in over 170 countries and regions, serving more than three billion people around the world. Huawei is committed to bringing digital to every person, home and organization for a fully connected, intelligent world.
In the transportation sector, including aviation, urban rail transit, railways, highways, water transport, ports and customs, urban traffic, and logistics, Huawei envisions "Pleasant Travel, Efficient Logistics, and Intelligent Wins." Through technological innovation and market layout, Huawei provides comprehensive intelligent solutions for the industry, driving the digital transformation and sustainable development of the transportation sector.
Currently, Huawei has served over 210 airports, airlines, and air traffic control centres, more than 300 urban rail transit lines in 70 cities, and over 180,000 kilometers of railways, more than 200,000 kilometers of road networks, and over 100 water transport and port industry customers.
In the past period, Huawei has been continuously advancing its All Intelligence Strategy by pursuing a series of innovations in the AI domain. These innovations support a vast range of AI models and applications across industries and are helping accelerate intelligence to reshape and upgrade industries.
Language models was launched in July 2023 and was already on Model 5.0 by the following year.
Huawei believes that open collaboration leads to mutual growth. Huawei is working with industry partners and developers to build open ecosystems in domains like HarmonyOS, openEuler, and openGauss. Together, Huawei aims to bring together all types of expertise across the industry to unleash ecosystem-based innovation and create a better digital and intelligent future.
Bringing AI into ITS
The relationship between Intelligent Transportation System (ITS) and digital telephony providers has morphed considerably over the last two decades.
This is a trend that has accelerated as later-generation cellular technology has been deployed, especially as 5G has addressed the low-latency communications requirements for safe operation of cooperative and autonomous vehicles.
For transportation and its ITS subset, the adoption of AI is a logical progression because, just as in other domains, if implemented and used correctly it offers significant potential.
“Fears over job insecurity, and being ‘replaced’ by machines, detract from the technology’s potential to be a true force for good”
Jacky Wang, vice president of Huawei’s Smart Transportation business unit, regards this process as being highly symbiotic. He starts by defining the role of data in transportation:
“Transportation is a traditional industry with different operating models for different modes such as air, sea, road, rail, metro, subway and pedestrian. All share a focus on the movement of people and goods. The essence of the industry is to move people and goods faster and with greater efficiencies, while maintaining and improving safety.
“For that to happen, we need data flows that connect people, goods and carriers so that transportation requirements and resources are better aligned. We need the necessary algorithms to do that and, in such a world I think that AI has a great deal of potential.”
Providing the backbone
There are, he feels, distinct roles — even if these are often intimately entwined.
“Huawei is an ICT infrastructure provider. We supply a unified digital and intelligence foundation. That means connectivity, whether via wireless or fixed networks, in combination with the cloud or data centres that provide the computing capabilities and storage, data governance and access to AI algorithms. We don’t define the algorithms unilaterally; we work with our partners in the various different domains such as transportation to make that happen.
“But if, as we do, you supply a unified foundation, how do you adapt it to each industry sector’s requirements? Cooperation with our partners is important, as is our ability to adapt our Large Model through the use of Large Natural Language and Computer Vision AI models.”
The use of different AI models enables Huawei to support different applications. The Computer Vision model, for example, lends itself readily to the inspection of rail wagon wheel sets. The ability of AI-based systems to rapidly identify images of wheels that are within safe operating limits means that human inspection of images is reduced by 95 percent — a dramatic efficiency gain.
Increasing accessibility
At the MWC Barcelona 2025 event in March, all of the key players in the mobile telephony market were keen to highlight their AI credentials. Huawei was able to demonstrate how its already well-established capabilities are evolving. These newer iterations support the company’s ambitions to increase AI’s capabilities while lowering the skill threshold for development and facilitating its quicker implementation, as Wang explains:
“This year, we have significantly strengthened our AI and big data product offer.
“The first stages of our AI effort were based on our Pangu model, especially in terms of predictive capabilities and the use of the Computer Vision model. Over the last several months, DeepSeek has become very popular and we already have some transportation-related references there, particularly in air and seaports, and in traffic policing.
“The top players in those sectors have typically used huge computer resources to achieve their goals. Now, they can just use a local knowledge library, adopt DeepSeek R1 or V3 and benefit from some custom assistance.
“For the moment if you want a high level creativity you need a person”
“An example is an airport in southern China, where eight people were tasked with collating 800 individual manuals which detail security procedures. No-one can memorise that amount of content but, by using DeepSeek in combination with Huawei’s FusionCube hypervisor/storage solution, security staff have been given on-demand access to all of the information that they need to know to be able to do their jobs.”
The speed of implementation is impressive, he notes.
“The DeepSeek/FusionCube is also being used in ports for container identification. Implementation requires a little budget, time and human resource allocation but it can produce appreciably positive results in a very short space of time. China Railways, for instance, replaced its entire customer assistance system in little more than a weekend.”
Evolution and acceptance
The application of the airport in Southern China mentioned above highlights how AI should be viewed as an enabler for the individual, rather than as a replacement.
Wang acknowledges people’s suspicions about AI’s increasing capabilities and that users’ intentions continue to cause concern. Fears over job insecurity, and being ‘replaced’ by machines, detract from the technology’s potential to be a true force for good. He says that it is important to recognise that AI has not matured as people expected.
“Innovation has been mainly in sister efforts to human tasks in the workplace. The focus has been on the tougher working environments, such as seaports and mining, or inspection of metro lines at night or in poor weather,” he explains.
“AI is great for fixed, regular, frequent tasks — repeatable stuff. It is much better than people for such things and it never gets tired. It can be used for more complex tasks but the cost point for such applications means that a human is still cheaper; we may reach a point where AI develops consciousness but for the moment if you want a high level creativity you need a person.”
To explain that further, he pushes us to look at how technology has already progressed.
“Go back a few hundred years, and most people were peasants. They lived and worked on the land. Nowadays, farming has become automated. We use machines and no-one complains about doing so — no-one is saying that they don’t want more automation, or that they don’t need it.
“At the same time, we didn’t gain more time. If anything, we’re even more busy.
“I don’t think that even if it becomes more capable AI will take people’s jobs. The technology needs to be arranged in tasks where it can help people. I believe that if we consider its use from a whole society perspective, then our ambition — our focus — should be to train the legacy people to do other things that have much greater intrinsic value,” he concludes.
Content produced in association with Huawei
