EN / 中文

Cross-platform Autonomous Haul Truck Fleet Collaboration at Zhonglian Runshi Open-Pit Coal Mine

by zhijiazuiqianyan·November 19, 2026

In September this year, an interesting change occurred during the expansion of unmanned haul trucks at the Zhonglian Runshi Open-Pit Coal Mine.

After 40 pure-electric unmanned haul trucks equipped with the EACON Intelligent Driving Intelligent Driving system were deployed, they did not open a new working area. Instead, they formed mixed fleets with existing unmanned haul trucks, sharing the same transport roads, loading zones, and dumping areas.

Public information indicates that the project broke down data silos between two autonomous driving platforms: EACON Intelligent Driving Intelligent Driving and Xinglu Intelligent Driving. This enabled unmanned haul trucks from different platforms and models to collaborate within the same mining site.

In the past, deploying autonomous driving in a mine typically meant one autonomous driving system corresponded to a batch of vehicles. Vehicles used the same platform, and positioning, task, and dispatching information flowed within the same system.

When another autonomous driving system joined, the emerging issues were no longer just about whether the new vehicles could drive autonomously.

The two systems possessed distinct vehicle-side algorithms and management logic, yet their vehicles had to enter the same transport roads, loading zones, and dumping areas. They needed to avoid mutual interference while operating around the same set of production tasks.

How exactly are different autonomous driving systems integrated into the same fleet?

01. How Do Vehicles from Different Platforms Enter the Same Operating System?

Cross-platform mixed deployment does not mean merging two autonomous driving systems into one.

Different vehicles still use their own perception solutions, positioning algorithms, and planning strategies. The real challenge is establishing a basic consensus on the on-site operating status among different systems once they enter the same mining area.

The most fundamental information is the vehicle's position.

If different systems adopt different spatial references for the same working area, or if there are discrepancies in the map information used, vehicles may form different judgments about each other's positions and road relationships.

Therefore, cross-platform collaboration first requires establishing a reliable information mapping mechanism, allowing different systems to interpret vehicle positions and operating information within the same working scenario.

However, position itself is not static data.

Vehicles are continuously moving, and mining roads may change due to construction, waste dumping, and other reasons. When a system receives vehicle information, it must also know when the data was generated, how long it took to transmit, and whether it remains valid at the current moment.

This is an easily overlooked issue in cross-platform collaboration.

Different systems must not only exchange information but also understand when this information remains trustworthy.

If a haul truck's position has already changed, but another system continues to make judgments based on outdated data, collaboration loses its meaning.

Thus, what needs to be established for different platforms to enter the same operating system is not a unified autonomous driving algorithm, but a foundational layer that allows different systems to share essential operating information.

Only after resolving what is happening in the mining area can subsequent vehicle coordination proceed.

02. Systems Must Also Understand What Vehicles Are Doing

If different platforms only exchange position, speed, and direction, they can only solve the most basic vehicle avoidance problem.

But autonomous mining transport is not simple road driving.

Starting from receiving a transport task, a haul truck goes through a series of steps: heading to the loading zone, waiting, loading, transporting, entering the dumping zone, unloading, and returning.

A haul truck decelerating could be waiting to enter the loading zone; encountering an anomaly on the road ahead; queuing to unload; or having just experienced a adjustment to its transport task.

Depending on the task stage, the collaborative strategy adopted by the system should differ.

Therefore, the information required for understanding in cross-platform mixed deployment can be divided into three layers.

The first layer is Motion Status, including vehicle position, speed, and direction, which describes how the vehicle is currently moving.

The second layer is Operational Status, such as whether the vehicle is empty or loaded, loading or transporting, waiting or unloading. This information tells the system where the vehicle stands in the overall production process.

The third layer is Task Intent, which includes where the vehicle plans to go next, whether the current task has changed, and if the transport route needs replanning.

The importance of these three layers varies.

Motion Status helps vehicles judge current safety distances; Operational Status allows the dispatching system to understand the entire transport flow; and Task Intent enables other vehicles and systems to predict future traffic relationships in advance. This is a key difference between autonomous mining and autonomous driving on open roads.

Vehicles on urban roads do not have a unified task center; other vehicles can mostly only infer their next behavior by observing trajectories.

In contrast, the operation of autonomous haul trucks is inherently part of the production process, and vehicle behavior can be understood in conjunction with task status.

Therefore, cross-platform mixed deployment cannot just allow different vehicles to see each other; it must enable different systems to understand what production stage the vehicle is in and what it might do next.

03. Who Coordinates the Entire Fleet While Vehicles Handle Driving?

The ability of different systems to share vehicle status is merely a prerequisite for mixed deployment, not completion of it.

When dozens of unmanned haul trucks from different platforms operate simultaneously, a critical question arises: who decides how tasks are allocated across the entire fleet?

Autonomous driving systems can enable a single vehicle to complete perception, planning, and control, but they cannot solve the transport efficiency problem of the entire mine.

If multiple vehicles head to the same loading point simultaneously, queuing may occur, reducing efficiency; if a dumping area becomes congested and subsequent vehicles continue to follow their original tasks, congestion will worsen.

At this point, the issue is no longer about how individual vehicles drive, but how the entire fleet operates. This involves two aspects: vehicle-side autonomy and the collaborative system.

Vehicle-side autonomous driving handles local and real-time decision-making.

Issues such as sudden obstacles appearing ahead or emergency road avoidance must be resolved by the vehicle locally through rapid perception and control. It is impossible to wait for remote systems to issue every driving command.

The higher-level collaborative system handles global task arrangement.

Decisions such as which vehicle goes to which loading point, which dumping area needs capacity adjustment, and which vehicles need reassignment after a transport road anomaly require integrating the production status of the entire mine.

The key to cross-platform mixed deployment lies in connecting these two aspects.

Different platforms do not need to use identical perception algorithms or completely consistent path planning methods. However, vehicles must be able to feed back their tasks and status to the collaborative system and receive dispatching information related to production tasks.

In simple terms, the vehicle side decides how to drive, while the collaborative system decides where to go and what to do. This division of labor allows different autonomous driving platforms to maintain their respective technical routes while participating in the same production system.

Therefore, what truly needs to be connected is not the driving decisions of all vehicles, but the information interface between vehicles and the entire mine production process.

04. Final Thoughts

When unmanned haul trucks from different systems enter the same mining area, their underlying technical routes may differ, but they must adhere to the same operational rules.

Sensor configuration, vehicle positioning, and path planning algorithms can be implemented independently by each platform.

What truly needs to be connected is the relationship between vehicles and the entire transport system: how vehicle status is expressed, how transport tasks are allocated and modified, how operational conflicts are coordinated, and how anomalies are transmitted to the dispatching system.

This effectively shifts the technological competition for unmanned haul trucks from single-vehicle intelligence to system collaboration, which is the opposite trend of passenger vehicle smart driving development.

In the past, companies primarily needed to prove whether their systems could enable a single haul truck to complete transport tasks safely and stably. In the mixed deployment phase, single-vehicle capability is just the foundation; the system must also handle task handoffs between different vehicles, allocation of operational resources, and resolution of running conflicts. Vehicles must not only complete transport independently but also be integrated into a larger production organization system.

This requires autonomous driving systems to retain their technical capabilities in perception, decision-making, and control, while also possessing the ability to exchange information with other platforms, accept task scheduling, and report operating status.

The interfaces and collaborative mechanisms between platforms will impact the transport efficiency of the entire mining area.