Chinese startup Yisen Power raises millions for AI dispatch brain for 8,000 autonomous delivery vehicles

As China's fleet of autonomous delivery vehicles surges past 50,000 units by early 2026, a new challenge emerges: how to coordinate thousands of vehicles across a single city. One startup is betting on a cloud-based AI dispatch system to solve it.

The Problem: Too Many Vehicles, Not Enough Coordination

China's urban delivery logistics market generates approximately 3 trillion yuan in annual revenue, with 25 to 30 million urban delivery vehicles on the road. Yet autonomous delivery vehicles currently account for only 1% of that fleet.

Meanwhile, total national logistics costs reached 19.5 trillion yuan in 2025, representing 13.9% of GDP. Last-mile delivery costs have surged from roughly 30% of total per-order costs a decade ago to over 60% today.

The trend toward unmanned delivery is clear, and the number of autonomous vehicles is growing rapidly. But with more vehicles comes a critical question: how do you match vehicles with orders, decide routing priorities, and integrate trunk lines with last-mile delivery?

Simply adding more vehicles won't solve the problem. What's needed is a unified system that can optimize vehicles, orders, battery levels, and road networks simultaneously.

The Solution: City-Level Cloud Dispatch

Yisen Power (易森动力) has developed a cloud-based AI dispatch hub called "Yisen Taichu" (易森太初). The system can take 1,000 autonomous delivery orders and calculate a globally optimal dispatch plan in just 8 seconds.

The company distinguishes between two dispatch models:

• Traditional dispatch systems (like Meituan or Didi): Match orders one-by-one to the nearest available vehicle. This works when the number of people and vehicles is fixed, but struggles with hundreds of autonomous vehicles and continuously rolling new orders across a city.

• City-level global dispatch: Considers all vehicles, orders, battery levels, routes, and time requirements simultaneously to produce an optimal overall plan. Yisen Taichu uses this approach.

In live demonstrations, the system presents two options for each shift: "Experience Priority" (more vehicles, higher on-time rate) and "Cost Priority" (fewer vehicles, less empty mileage, shorter total distance). The trade-off between vehicle utilization and punctuality is made transparent for human decision-makers.

Results show that with the same batch of orders, the vehicle-to-order ratio improves from 1:2 to 1:4, the number of vehicles deployed decreases, and route planning time drops from tens of minutes manually to 3 seconds.

The Business Model: Becoming the "Chain Leader"

The value of a dispatch hub lies in the cost structure of autonomous vehicles. A single vehicle running only 2-3 trips per day fails to amortize its cost. But when integrated into a dispatch network, 30,000 to 40,000 autonomous vehicles in a city could handle the work currently done by 300,000 manned vehicles.

This has driven a shift across the industry from "selling vehicles" to "selling delivery capacity." Companies like Jiushi Intelligence (九识智能) now offer vehicle leasing and unmanned delivery franchise models, charging by the day or hour.

The dispatch hub is the core engine that integrates fragmented delivery capacity. It determines utilization rates and who becomes the "chain leader" — the entity that coordinates an entire city's autonomous delivery network.

On September 20, Yisen Power announced a strategic partnership with SAIC Maxus (vehicles), Zhuoyu Technology (卓驭科技, intelligent driving), and CATL's Times Electric Service (时代电服, batteries and swapping). The consortium plans to deploy 50,000 battery-swappable autonomous logistics vehicles — reportedly China's first deployment at the 10,000-vehicle scale.

Yisen Power contributes the "operations + dispatch + settlement" platform and acts as the chain leader that brings the partners together.

"We're not building a full-stack intelligent driving company," said Wang Xuan, founder of Yisen Power. "Intelligent driving is already a commodity. We want to be the chain leader, aggregating the best partners across the industry."

The business model works as a flywheel: customers pay for guaranteed delivery, vehicle owners contribute vehicles for profit-sharing, operators handle local delivery, and the platform coordinates dispatch and settlement. Each completed order feeds data back to the platform — more data means better dispatch, better dispatch means higher utilization, higher utilization means lower unit costs, which attracts more customers and capacity.

In less than a year since founding, Yisen Power has:

• Operated over 50 vehicles
• Secured over 8,000 units in意向 orders
• Expanded to 9 cities
• Counts Hebei Express Group among its clients

The Competitive Landscape

The dispatch hub market is attracting multiple players:

• Internet platforms (Meituan, Didi) already have order entry points and data. As autonomous fleets scale, they could extend their dispatch systems to urban delivery.

• Automakers and intelligent driving companies (Dongfeng, Changan, GAC) are entering the autonomous vehicle space but may lack expertise in operations, revenue management, and fleet coordination.

• Established players like Jiushi and New Stone (新石器) have formed a "duopoly," commanding nearly 90% of the market share.

• JD Logistics has announced a target of deploying 1 million autonomous vehicles within five years.

As leading players accumulate larger fleets, the question of "who coordinates the middle" becomes increasingly critical — will platforms build their own dispatch systems, or outsource to specialized "chain leader" companies?

Future Outlook

The dispatch business's potential depends on how many scenarios autonomous vehicles can penetrate. The most immediate applications include express delivery, medical logistics, convenience stores, and hotel linen services — areas where autonomous vehicles can achieve significant near-term substitution.

The key bottleneck is the "last 50 meters" — delivering to upper floors, picking up and moving goods. This will require mature embodied robots, representing a longer-term variable.

Wang Xuan predicts that China's autonomous logistics vehicle fleet could reach 1.5 million units within five years.

"Three trillion yuan is an abstract number," he said. "For context, Didi's ride-hailing market is a 400 to 600 billion yuan sector. In a 3 trillion yuan market, we won't see a red ocean for 5 to 10 years. Everyone has enough time."

This article does not constitute investment advice.