Traditional urban transport monitoring methods, being labor-intensive and slow, are unable to provide the real-time data necessary for fast decision-making. Manual inspections take weeks, delaying the identification of critical issues like road wear or structural damage. Existing UAV systems, while useful, face limitations such as short battery life, slow data processing, and connectivity challenges, making them inadequate for dense urban environments.

In response to these challenges, AAI Labs is helping to develop a 5G-based UAV base station system. This fully autonomous system, equipped with AI-driven photogrammetry and intelligent image recognition, will provide  real-time, accurate monitoring of urban transport networks.

Solution

AI-powered UAVs‍

The UAVs are equipped with AI-powered systems for object detection and tracking, including a pipeline based on the YOLOv10 model. This allows for the detection and tracking of persons and vehicles in both normal and thermal drone footage. Tracking performance is measured through metrics such as the MOTA score (for multi-object tracking accuracy) and the IDF1 score (for identifying objects across frames). Detection accuracy is quantified by mAP(50), which has a performance of 66-79%. The AI system enables high-resolution data capture using photogrammetry and advanced image recognition, which can instantly identify infrastructure issues like road damage or traffic congestion.

Real-time data transmission via 5G‍

The integration of 5G technology ensures high-speed, low-latency communication between UAVs and ground stations. This enables real-time data transmission and processing, allowing for faster decision-making by urban planners and engineers. By using 5G capabilities, the system bypasses the connectivity issues that often plague conventional UAV systems.

Automated battery swapping

‍Another limitation in traditional UAV operations is the short battery life. To address this, our solution introduces an automated battery-swapping system, which allows UAVs to operate continuously without the need for manual battery changes. This feature is particularly critical in enabling non-stop surveillance in dense urban environments. 

Georeferencing for accurate location data

‍Our system also includes a georeferencing algorithm that calculates GPS coordinates with high accuracy using the drone's known location and ground sampling distance. This feature provides the precision needed for medium-accuracy applications, which is essential for mapping and monitoring infrastructure.

Flight controllers for precision and stability

‍To ensure stable flight and precise control, the system employs advanced flight control mechanisms, including Proportional-Integral-Derivative (PID) and Model Predictive Control (MPC) controllers. PID controllers are optimized through differential evolution to maintain desired altitude and minimize movement overshoot, while MPC  predicts future behavior to follow predefined flight trajectories closely. 

Applications and impact of the solution

• UAVs equipped with intelligent vision systems provide continuous scans of roads, sidewalks, and parking areas. This real-time data enables city planners to detect early signs of infrastructure wear, leading to proactive repairs, cost reductions, and enhanced public safety.

• Drones monitor air quality and assess environmental conditions, identifying hazards like overgrown vegetation obstructing roads. This allows cities to take preventive actions, improving urban safety and environmental sustainability.

• Real-time aerial surveillance of traffic patterns allows cities to dynamically adjust traffic signals, optimizing vehicle flow and reducing congestion.

The system significantly shortens infrastructure inspection times from 6-8 weeks to just hours, accelerating repairs, lowering costs, and improving safety. Continuous, automated monitoring through battery-swapping UAVs ensures uninterrupted operations, setting a new standard for urban surveillance technology. Solution’s scalability means that cities of any size can deploy this technology, making urban environments smarter, safer, and more sustainable in the long run.

Partner