Didi's "own child" gave Elon Musk a lesson.

The long - dormant autonomous driving track has been awakened by a sudden huge amount of financing.

Recently, KargoBot announced the completion of a Series B financing round exceeding $100 million. This is also the second financing round exceeding $100 million in the autonomous driving field since entering 2026.

Who is KargoBot?

Looking at its resume, it is almost the "second - generation scion" in the autonomous driving industry: It was spun off from DiDi's autonomous driving division, helmed by former CTO Wei Junqing. Behind it stands Horizon Robotics, one of the four giants in intelligent driving, and the state - owned investment in Ordos has been continuously increasing its stake since the angel round.

More importantly, it was founded at the peak of the last autonomous driving boom in 2021. This "rich second - generation" born with a silver spoon in its mouth could have competed with the Robotaxi (driverless ride - hailing) giants for the dominance of the streets in first - and second - tier cities, but it actually chose a complete "rural movement".

While its peers were still struggling with takeaway electric scooters and unpredictable long - tail scenarios at complex intersections in Beijing, Shanghai, Guangzhou, and Shenzhen, KargoBot had turned the steering wheel and plunged into the desolate trunk roads in the northwest, targeting the Robotruck (driverless freight truck) for coal and ore transportation, and earned its first pot of gold.

Behind this is a cruel choice about survival.

Today's autonomous driving is experiencing a similar disillusionment and rebirth as humanoid robots. The technology story is still appealing, but the patience of capital has been exhausted. The focus of the market has returned to the origin: Can it run stably, run far, and run profitably?

Behind this $100 million financing is a overtaking of Robotruck over Robotaxi on a curve.

"Carpooling" in Trunk Logistics

The birth of KargoBot seems like a reverse bet.

In February 2021, at the peak of the last wave of the autonomous driving industry boom, DiDi spun off KargoBot from its subsidiary. At that time, DiDi had accumulated a large amount of Robotaxi data in cities such as Guangzhou and Shanghai. However, it was precisely this data that made DiDi's senior management sense a hint of unease:

The complex road conditions in urban areas, the public's trust deficit in driverless technology, and the bottomless profit black hole have made the short - term commercialization prospects of Robotaxi extremely unclear.

So, KargoBot was sent to the more "rough" battlefield of trunk logistics. This decision seemed even a bit tragic at that time, because it meant moving from the constant - temperature office buildings in Lujiazui to the heavy - truck repair shops in Inner Mongolia and Xinjiang.

China's trunk logistics is a vast but inefficient wasteland. Nearly ten million heavy trucks carry about 70% of the country's freight volume, accounting for nearly 5% of the GDP, but are struggling with a meager profit margin of less than 1% under the compliant freight rates. The high empty - running rate of up to 40% and the rising labor costs have made this business almost like "bending down to pick up coins", and the demand for cost reduction and efficiency improvement is extremely urgent.

The core team sent by DiDi is the key to ensuring the project's implementation. CEO Wei Junqing was one of the earliest doctoral students in the Carnegie Mellon Autonomous Driving Laboratory. After serving as the global engineering vice - president of Aptiv, he joined DiDi in 2019 as the CTO of autonomous driving and officially took over KargoBot in 2021, leading the implementation path of autonomous driving in trunk logistics.

Since its independent operation in 2023, KargoBot has delivered an impressive report card:

As of the end of 2025, about 400 L4 - level autonomous trucks have been in regular operation on trunk roads in more than ten provinces and cities, including Inner Mongolia, Shaanxi, and Xinjiang, with a cumulative mileage of over 35 million kilometers and a freight volume of 1.2 billion ton - kilometers. In Inner Mongolia, KargoBot holds more than 50 driverless operation licenses, accounting for 90% of the national total.

On the 300 - kilometer route from Ordos to Baotou, a mixed - formation truck can complete 4 trips of transportation per day. Originally, the single - trip freight was about 15,000 yuan. After optimization through the autonomous driving solution, the daily net increase in efficiency exceeds 12,000 yuan, and the annual gross profit can increase by nearly 4 million yuan; the annual labor cost per truck has decreased by 83%, and the entire fleet saves about 330 million yuan per year. This has made KargoBot one of the few real cases of commercial implementation in the industry.

Behind the achievements lies a technical compromise of "carpooling": a hybrid intelligent formation of "a manned lead vehicle and unmanned following vehicles".

This solution was once despised by the "fully driverless" extremists, who thought it avoided the long - tail problems of AI operating alone. However, in Wei Junqing's view, this is an extremely pragmatic approach based on human nature: leaving the complex human - related matters such as loading and unloading goods and passing through toll booths to the drivers, and entrusting the monotonous and high - pressure long - distance trunk routes to AI.

This logic of "humans do what they are good at, and machines do what they are good at" has directly tripled to sextupled the gross profit per truck. At the same time, it takes into account compliance and implementation without waiting for the full maturity of single - vehicle intelligence. It can be adapted to the bulk commodity transportation scenario, avoiding the policy restrictions of fully driverless operation, and is more likely to be recognized by regulators and logistics enterprises.

Wei Junqing believes that even if the driverless technology matures in the future, the formation mode of three to four vehicles will still be the most efficient solution for bulk transportation. It is estimated that the mixed - formation mode will account for about 50% of the autonomous driving market in trunk logistics and become the mainstream operation mode in the long term.

Horizon Robotics Gets Involved

In this round of financing, the role of the leading investor, Horizon Robotics, is worthy of attention: It has been upgraded from a chip supplier to a strategic shareholder.

In the past few years, the autonomous driving industry has experienced obvious fluctuations in temperature: the financing peak was nearly 100 billion in 2021, plummeting to 24 billion in 2022 and further dropping in 2023. TuSimple delisted, Waymo's truck project was postponed, and Alibaba's "Big Donkey" was reassigned. Capital has gradually understood that only those who can implement their technologies are qualified to survive.

Against this background, the logic for Horizon Robotics to further increase its investment in KargoBot is not complicated.

KargoBot has achieved large - scale operation of L4 - level vehicles in the trunk logistics scenario, with a cumulative mileage of over 35 million kilometers, and its commercialization indicators are quantifiable.

Horizon Robotics' Journey® 6P computing platform will further promote the localization of hardware and optimize the cost structure; its accumulation in vehicle manufacturers and parts supply chains will also support KargoBot's network expansion; as a result, the commercialization efficiency may be further improved, and the plan is to deploy a thousand autonomous trucks in 2026.

Horizon Robotics' involvement also means a deep binding of technology implementation and business logic.

The KargoBot Space jointly launched by KargoBot and Horizon Robotics adopts a driver - cab - less design, increasing the cargo compartment space by 25%, the effective load by 10%, and the annual revenue per truck by 250,000 to 400,000 yuan. These data show that technological innovation is not just a simple concept but can be directly transformed into economic value.

The ecological synergy effect is gradually emerging. Hongqi's investment provides vehicle resources, Dongfang Jialian optimizes the supply chain, and local governments ensure road rights, making the operation efficiency of the northwest fleet about 30% higher than that of its peers. With the promotion of the mixed - formation mode, 6 new cross - provincial transportation routes were added in 2025, with a daily freight volume of more than 800 tons, and the efficiency of trunk logistics almost doubled.

Wider industry signals are also clearly visible. In October 2024, Neolix completed a $600 million Series D financing, and in February 2026, Jiushi Intelligence completed a financing of over $300 million. Together with KargoBot, this indicates that capital has set its sights on the Robotruck and driverless logistics tracks.

Compared with Robotaxi, trunk logistics has greater short - term commercialization potential and is becoming the core engine of the industry's recovery.

Get Ahead of Robotaxi

The rise of KargoBot and the preference of capital for the Robotruck track essentially reflect the differences between the Robotruck and Robotaxi tracks. Whether from the perspective of implementation difficulty or commercialization potential, Robotruck is obviously more likely to achieve large - scale operation and can quickly realize the value of technology.

Taking Tesla's Robotaxi as an example, since the pilot in Austin in June 2025, as of February 2026, only 42 vehicles were actually in operation, and the peak availability rate was less than 20%. A large number of vehicles were idle or under maintenance, and the conversion of 8.2 billion miles of supervised mileage into commercial operation mileage was negligible. The safety risk is still high: there were 14 collision accidents, and based on the known mileage, the accident rate is about 4 times that of human driving.

In contrast, KargoBot's Robotrucks operate in the trunk logistics scenario, with fixed routes and low traffic complexity. The mixed - formation mode of one lead vehicle leading 2 to 6 L4 driverless vehicles avoids the long - tail problems of single vehicles and reduces labor costs at the same time. Its safety performance has exceeded that of human driving by 5 times, supporting unmanned ETC lane - opening on highways and realizing full - process unmanned operations.

In terms of business logic, the profit path of Robotruck is clearer. Robotaxi targets C - end users. Although it has technological accumulation, the vehicle utilization rate and order volume are low, and the cost is high, making it difficult to make a profit.

KargoBot targets B - end logistics enterprises, with the core goal of cost reduction and efficiency improvement. Under the mixed - formation mode, the annual income of each autonomous truck is about 1 million yuan, and the driver cost accounts for about one - third. After installing a 90,000 - yuan - level autonomous driving kit and amortizing it over 5 years, the annual labor cost can be saved by about 200,000 yuan, and the pay - back period in some scenarios is only 6 months.

Wei Junqing said that when the fleet size exceeds 3,000 vehicles, the company can achieve self - sufficiency and officially cross the profit inflection point.

Policy and industrial synergy further reduce the resistance to implementation. Logistics - heavy provinces such as Inner Mongolia and Shaanxi have opened up road rights and provided policy support. Coupled with the cooperation between local governments and logistics enterprises, KargoBot can quickly deploy vehicles. The industrial chain synergy is complete, which is conducive to technology implementation and scale expansion. Robotaxi has higher coordination costs in terms of policy, road rights, and urban management, and its implementation pace lags behind.

Of course, Robotruck still faces challenges, including insufficient coordination of cross - provincial road rights, adaptability to complex roads, and room for hardware cost optimization.

However, compared with Robotaxi, its implementation path is clearer, and its commercialization potential is easier to realize. The judgments of capital and investors also confirm this: Robotruck has become a trillion - level track, and KargoBot has entered the stage of large - scale development with the verification of its algorithms, data, and mixed - formation mode.

With the financing completed and the technology, policy, and capital paths in place, the "scale effect" of autonomous driving in trunk logistics is approaching the critical point. But the real battle is not about algorithms or sensors, but about replication ability, cost control, and operation management.

Can the Ordos model be replicated ten times, a hundred times? Can the number of 400 vehicles be expanded to more than 1,000? Can the data of 1.2 billion ton - kilometers continue to grow and be transformed into a profitable fleet? These are the tough battles that KargoBot will face in the next stage.

If successful, it will not only leverage the trillion - level trunk logistics market but also may redefine the survival logic of autonomous driving in China.

Reference Materials

Leifeng.com "KargoBot Completes Over $100 Million in Financing, Wei Junqing: Let Driverless Heavy Trucks Cross the Desert Gobi"

21st Century Business Herald "Spun off from DiDi, Led by Horizon Robotics! KargoBot Gets $100 Million in Financing"

Cailian Press "Driverless Logistics Sees Another Big Financing! KargoBot's Cumulative Financing Reaches 2 Billion"

China Newsweek "Will No One Be Driving Soon?"

DoNews "KargoBot Achieves the World's Largest Commercial Operation of L4 Driverless Heavy Trucks in Ordos"

This article is from the WeChat public account "Golden Horn Finance", author: Chester, published by 36Kr with authorization.