As drone technology rapidly advances and policy documents promoting drone applications in the highway sector are intensively released, unmanned aerial vehicles (UAVs) are increasingly deployed across construction, management, maintenance, operation, and service scenarios of expressways. Drones offer advantages such as wide monitoring range, flexibility, unrestricted spatial access, and rapid on-site arrival. They can quickly reach incident locations, acquire real-time aerial video and images, and operate efficiently even in complex and hazardous environments like bridges and slopes. This makes them an effective tool for high-quality development of the highway industry. However, issues such as flight safety risks, insufficient synergy with existing operations, lack of shared utilization, and low efficiency of manual operations have become increasingly prominent. Ensuring the safety of drone applications has emerged as a major challenge for the industry.
Current Status of Highway Drone Applications
National Policies Intensify Drone Regulation
On January 1, 2024, the “Interim Regulations on the Management of Unmanned Aircraft Flight” (hereinafter referred to as the “Regulations”) jointly issued by the State Council and the Central Military Commission officially came into effect. The Regulations explicitly designate the airspace above highways as controlled airspace, prohibiting unmanned aircraft flight activities without approval from air traffic management authorities. Additionally, the Regulations stipulate requirements such as mandatory drone registration, insurance, a permissible flight altitude below 120 meters for unmanned aircraft, emergency response plans, and reporting of safety incidents to air traffic control within 24 hours. Operators of drones other than micro and light models must hold a flight license, while micro, light, and small drones may fly without declaration in permissible zones; all other scenarios require prior flight approval. The Regulations represent China’s first civil unmanned aircraft law, ending the long period of regulatory vacuum. Under the trend of increasingly strict national drone regulation, ensuring safe and compliant drone use has become a prerequisite for effective deployment.
Existing Problems in Highway Drone Applications
1. Lack of unified planning: Different highway sections conduct independent drone operations using various drone models and manufacturer platforms, leading to fragmented application scenarios. Superior management units cannot grasp the overall drone usage situation, creating difficulties in coordination and unified management. Especially when equipment resources are limited and flight tasks are numerous, it is necessary to arrange flight times and efficiently dispatch pilots and equipment.
2. Safety hazards: Drone operations mainly rely on manual on-site control by pilots, whose skill levels determine operational effectiveness. This results in low overall efficiency, inability to guarantee route accuracy and full area coverage, and limited visual range for pilots. When large-scale operations are required, there is a risk of collision with obstacles. Furthermore, data collected by drones is often transmitted via offline manual copying, posing risks of sensitive data leakage.
3. Lack of shared utilization: Each highway section procures and operates drones independently, leading to low resource utilization. Departments such as maintenance, road administration, and monitoring within the same section collect data separately, preventing data sharing. Each section develops drone applications independently, resulting in redundant construction, lack of application sharing, and poor iteration capability.
New Requirements for Highway Drone Applications
Currently, highway drone applications are in a development phase, with low alignment between drone technology and industry needs. There is an urgent need for systematic, standardized, and standardized management platforms to accelerate deep integration with industry demands:
- Supervision: Monitor all elements including drone equipment assets, flight airspace, pilot qualifications, flight status, and collected data to ensure safety and comprehensive awareness.
- Collaboration: Integrate drone capabilities into business systems according to workflows, allowing operators to seamlessly use drones without switching between systems, ensuring business continuity.
- Sharing: Real-time transmission and sharing of drone-collected data to business systems to meet live streaming and data application needs; concurrently, through comprehensive monitoring of drone usage, coordinate sharing across sections and departments to improve utilization.
- Automation: Enable scheduled and fixed-point automated flights to solve the high difficulty and low efficiency of manual piloting, enhancing operational safety.
Development of Domestic and International Drone Management Systems
With the advancement of drone technology, various drone management systems have been developed both domestically and internationally. Domestic systems include Zhongke Yuntu’s “E-Fei Cloud,” Fuya Intelligent’s “MindView,” DJI’s “DJI Smart,” and Autel Robotics’ “Autel Commander.” International systems include Dronepoint’s “UgCS” and Micridrones’ “MdCockpit.” However, most of these systems are either industry-general or tailored to specific scenarios, exhibiting limitations when applied to highways, as summarized in the table below.
| System | Disadvantages |
|---|---|
| E-Fei Cloud | Complex functions, not user-friendly |
| MindView | PC-only software, inconvenient for field pilots; cannot meet on-site management needs |
| DJI Smart | Complex route planning, high professional requirements; standalone system with no business collaboration or data sharing |
| Autel Commander | Compatible with only a few drone models, insufficient interoperability; standalone system, no collaboration, no data sharing |
| UgCS | Compatible with limited drone models, insufficient interoperability; standalone, no collaboration, no data sharing |
| MdCockpit | No route planning support; standalone, no business collaboration |
The above analysis reveals that traditional management systems lack the ability to seamlessly integrate with highway business processes, enforce drone regulation, and support automated operations across multiple departments. This gap necessitates the development of a dedicated highway drone centralized management system.
Innovation of the Proposed System
The proposed highway drone centralized management system incorporates several innovations:
- Integrated drone management platform: Unifies access to all highway drones and provides on-demand services including route planning, task planning, automated operations, flight monitoring, remote control, and data management, enabling rapid construction of drone applications while ensuring compliance with drone regulation.
- Flexible dispatch and command: Provides on-duty response capability, supports remote drone dispatch and automated operations without requiring personnel to be on-site, enhancing flexibility and efficiency.
- Data collaboration and sharing: Real-time sharing of drone-collected data with business systems to maximize data synergy, avoid duplicate collection and processing, and improve work efficiency.
- Multi-scenario application services: Deep integration of drones with business systems allows seamless invocation of drone services, enabling emergency command, routine patrol, and other scenarios.
System Architecture
The centralized management system adopts a SaaS-based design, consisting of perception layer, data layer, service layer, capability support layer, presentation layer, and system integration layer, which provides capabilities to application layer business systems as needed. The overall architecture is depicted below. (Please note that the figure is inserted at the end of this paragraph for reference.)
The layers are described as follows:
- Perception layer: Integrates drones, airports, and payloads to ensure accurate and real-time data collection.
- Data layer: Centrally stores video and images collected by drones, equipment basic data, route data, flight monitoring data, meteorological data, and no-fly zone data. Other business systems can access these data via service interfaces.
- Capability support layer: Includes high-precision maps, 2D maps, and video streaming center, providing geographic and video processing capabilities.
- Service layer: Implements functions such as route management, task management, drone monitoring, live streaming, remote control, automated operations, data management, and device management. Business systems can integrate these functions to avoid redundant development.
- System integration layer: Supports integration with other business systems via data service integration, drone management service integration, and permission control.
- Presentation layer: Develops PC and mobile applications with intuitive user experience.
- Application layer: Supports building drone applications for emergency rescue, maintenance patrol, road administration patrol, etc.
System Functions
The drone centralized management system consists of two subsystems: a PC-based management system and a mobile flight control App. The relationship is that business systems invoke services to plan routes and tasks, the management system receives instructions, controls drones for automated or manual operations, and shares flight information and collected data in real time. The App serves as the main tool for on-site manual piloting, enabling communication and data synchronization between the drone and the management system.
Management System (PC)
The functions of the PC management system include:
| Function | Description |
|---|---|
| Route Management | Support route planning, import, and viewing for centralized control of drone routes, ensuring compliance and safety. |
| Task Management | Manage task plans (automated or pilot tasks), set execution time, route, drone equipment; view historical tasks and replay flight trajectories. |
| Drone Monitoring | Real-time map-based monitoring of drone distribution, flight status, abnormal weather alerts, and no-fly zone warnings. |
| Drone Live Streaming | Real-time video live streaming during flight, supporting multiple simultaneous streams. |
| Remote Control | Remotely control drones and payloads for operations such as adjusting direction, speed, height, zoom, and capturing images. |
| Data Management | Ingest, store, view, and analyze data from drones and payloads; ensure data security and sharing. |
| Device Management | Support access of multi-model drones, airports, and payloads; view device info and health monitoring. |
| No-fly Zone Management | Import no-fly zone data, define custom zones, and view zones to ensure flights stay in safe areas. |
| Weather Data Management | Real-time weather data access and display. |
| Pilot Management | Manage pilot information, view certification certificates and task execution history, ensuring pilot qualifications comply with drone regulation. |
Flight Control App (Mobile)
The App functions include:
| Function | Description |
|---|---|
| Drone Connection | Connect to drone for task execution. |
| Task Management | View pending tasks (execution time, route) and execute them. |
| Flight Monitoring | Map-based monitoring of flight status, task progress, no-fly zone warnings, and weather alerts. |
| Video Live Streaming | View real-time video during flight. |
| Drone Control | Manual remote control: adjust speed, height, hover, take photo, record video, return home. |
| Data Management | View and upload collected data; monitor upload status. |
| No-fly Zone Query | View no-fly zones and restrictions on map. |
SaaS-based Integration with Business Systems
The centralized management system provides SaaS capabilities, offering atomic-level service interfaces for all functions. Business systems can build multi-scenario drone applications via two integration methods:
- Interface integration: Call a series of APIs to obtain drone-collected data or use management functions.
- UI integration: Embed SaaS pages for drone management functions without redeveloping interfaces.
Key Technologies
Automated Drone Operations
Automated operations aim to achieve scheduled autonomous flight, precise takeoff/landing, and real-time data transmission, improving efficiency and reducing safety risks from manual piloting. The system uses a task scheduling mechanism: based on predefined routes, execution time, and target (airport or pilot), it generates a drone schedule and performs conflict detection. For airport-based tasks, the system sends task and route information to the airport, which controls the drone to fly autonomously along the route, with data automatically transmitted back. For pilot-based tasks, the pilot receives the task via the App; upon starting, the drone automatically follows the route while data is streamed back. This approach minimizes manual intervention and adheres to drone regulation by ensuring pre-approved routes and controlled airspace.
Real-time Drone Monitoring
During flight, real-time monitoring of position, status, and task progress is essential. The system obtains drone latitude/longitude data via onboard sensors and API, dynamically displays the current position on a GIS map, and draws the actual flight path. By comparing with the planned route, task progress is tracked and any deviation triggers alerts. The real-time flight parameters include:
$$
\boldsymbol{P}(t) = (x(t), y(t), z(t))^{\text{T}}
$$
$$
\boldsymbol{v}(t) = \frac{d\boldsymbol{P}(t)}{dt} \approx \frac{\boldsymbol{P}(t+\Delta t) – \boldsymbol{P}(t-\Delta t)}{2\Delta t}
$$
$$
\text{Altitude deviation} = |z(t) – z_{\text{planned}}(t)|
$$
If the deviation exceeds a threshold (e.g., 5 meters), the system alerts the operator and records the event for compliance with drone regulation. After flight, the historical data is stored for replay.
Multi-channel Remote Video Transmission
Real-time video from multiple drones is transmitted to the management system using streaming technology. The workflow is as follows:
- Drone captures video.
- Video is transmitted to the streaming server:
- For airport-controlled drones: drone → airport → RTMP → streaming server.
- For App-controlled drones: drone → remote controller → App → RTMP → streaming server.
- The management system uses HTTP-FLV protocol to pull streams from the streaming server, enabling multi-channel live viewing. Video is also recorded for later analysis.
The video data rate can be modeled as:
$$
R = \frac{S}{T} \quad \text{where } S \text{ is total data size (bits) and } T \text{ is duration (s)}
$$
To ensure low latency and robustness, adaptive bitrate streaming and error correction techniques are employed. This real-time capability allows business systems to instantly access drone feeds, supporting emergency response while respecting drone regulation by keeping data secure and traceable.
Remote Control of Drones
During flight, operators may need to remotely control the drone or payload (e.g., adjust gimbal, zoom, take photos). Two flight modes are implemented with seamless switching:
- Normal monitoring mode: Real-time status (position, battery, speed, height, signal strength) is displayed via drone SDK, providing full situational awareness.
- Remote control mode: Operators can remotely command flight adjustments (yaw, pitch, roll), payload control, return-to-home, or pause. A user authentication mechanism locks control rights to prevent multi-user conflicts.
The control command latency is critical and must be bounded:
$$
\text{Latency} = t_{\text{command}} + t_{\text{transmission}} + t_{\text{drone response}} < \tau_{\text{max}}
$$
Typical requirements set \(\tau_{\text{max}} \leq 500\) ms for safe operation. This ensures that remote control does not violate drone regulation by maintaining positive control authority at all times.
Application Case Study
The drone centralized management system has been deployed to provide drone management services for the high-definition digital map project of Guangdong Transportation Group. By connecting multiple drones to the system, real-time position monitoring, flight status tracking, video live streaming, and historical data replay are achieved based on high-precision digital map layers. This has enabled the construction of an “air-ground integrated” monitoring service, supporting diverse highway scenarios such as emergency response, routine patrol, and incident management, all while ensuring compliance with drone regulation through centralized route approval, no-fly zone enforcement, and pilot credential verification.
Conclusion
The highway drone centralized management system effectively addresses the pain points encountered in drone applications, providing a solid technological foundation for diverse drone needs. By integrating drone capabilities into business systems and sharing data in real time, it enables efficient construction of drone applications across the entire highway lifecycle. Moreover, it ensures that all operations adhere to the stringent requirements of drone regulation, thereby promoting safe, compliant, and high-value drone utilization. With the increasing regulatory focus on unmanned aircraft, such a system is poised to become an indispensable component of intelligent highway management, driving digital transformation while safeguarding airspace security.