Warehouses are a typical indoor environment where drones progress from simply "being able to fly" to "being able to perform tasks." Unlike open environments, warehouses feature densely packed racks, narrow aisles, and complex cargo stacking patterns. GPS signals may be weak or unavailable, placing greater demands on drone localization, obstacle avoidance, and path planning.

At the same time, warehouse management is concerned not only with "where the cargo is," but also with "how much space it occupies." Traditional manual inspections and volume calculations are time-consuming and make it difficult to obtain warehouse-space data frequently and continuously.

If a drone can autonomously inspect within predefined warehouse boundaries, scan cargo and make rough volume estimates during flight, and simultaneously save inspection images and measurement data, it can form a more complete closed-loop digital warehouse inspection workflow.

How can a drone navigate autonomously through a warehouse, avoid obstacles, scan cargo, and ultimately generate a readable inspection report automatically?

This is precisely the focus of this installment's innovation topic: Autonomous Warehouse Inspection and Cargo Volume Estimation, which explores the autonomous operating capabilities of drones in smart warehousing scenarios.

01

Challenge Topic Overview

Autonomous Warehouse Inspection and Cargo Volume Estimation

This topic focuses on indoor warehouse inspection and warehouse-capacity sensing, exploring methods for a single drone to autonomously complete warehouse inspections, cargo scans, rough volume estimates, and data archiving. The core tasks are listed below, with the ultimate goal of transitioning from "manual inspection and manual tallying" to "autonomous drone inspection, automated estimation, and report generation."

Based on predefined warehouse dimensions and inspection areas, autonomously plan inspection routes and complete flight, turning, and real-time obstacle avoidance inside the warehouse;

Scan cargo observed during the inspection from multiple angles, acquire imagery or spatial data, and roughly estimate the volume occupied by the cargo;

Organize cargo-volume estimate data, corresponding inspection images, and necessary inspection records, then automatically generate a structured inspection report.

The drone must have the following capabilities:

Autonomous localization and mapping;

Environmental perception and obstacle avoidance;

Autonomous inspection and coverage path planning;

Cargo scanning and 3D data acquisition;

Approximate cargo volume estimation;

Data logging, image archiving, and report generation. 

Challenge Objectives

Five technical focus areas for an end-to-end warehouse inspection, estimation, and data workflow.

Autonomous Localization and Mapping

GPS signals are weak or unavailable inside warehouses. The drone therefore needs to continuously estimate its position and attitude by combining visual and inertial data, while simultaneously building a map of the warehouse environment. The localization results must support stable flight and also be used to record inspection routes and mark cargo locations.

Technical Focus:

Visual Localization / Inertial Navigation / SLAM

Route Planning and Autonomous Obstacle Avoidance

The drone must use sensors to obtain information about its surroundings, understand the spatial structure, and identify building debris, trees, slopes, or temporary obstacles, providing a basis for obstacle avoidance and safe passage. Perception results must also be transmitted to the planning module promptly to avoid situations in which an obstacle is detected too late for the drone to evade it.

Technical Focus:

Coverage Path Planning / Obstacle Detection / Waypoint Management

Object Recognition and Multi-View Scanning

After reaching the designated area, the drone identifies the cargo to be measured and determines the scan area. It then captures images, depth data, or point clouds from different distances and angles, providing complete and valid data for subsequent 3D reconstruction and volume estimation.

Technical Focus:

Object Recognition / Multi-View Data Acquisition / Depth Sensing

Scale Calibration and Volume Estimation

Camera calibration or a reference object of known dimensions is used to establish the relationship between scan data and real-world dimensions. The multi-view data are then registered, background elements such as the floor and racks are removed, the cargo region is extracted, and its actual volume is estimated.

Technical Focus:

Camera Calibration / Point Cloud Registration / Region Segmentation / Volume Estimation

Data Archiving and Report Generation

Cargo locations, on-site images, scan data, volume estimates, and inspection records are linked and stored. Reports are then generated by task ID or cargo ID to facilitate subsequent viewing, verification, and traceability.

Technical Focus:

Data Association / Image Archiving / Results Visualization / Report Generation

Application Scenarios

Autonomous warehouse inspection and cargo volume estimation are not only a technical challenge for drones; they also support multiple real-world applications in warehouse digitalization.

Smart Warehousing

Helps automate routine warehouse inspections and reduce repetitive manual checks.

Inventory and Warehouse Capacity Assessment

Quickly determines the space occupied by cargo and the warehouse's remaining capacity through cargo scanning and volume estimation.

Logistics Scheduling

Provides data to support inbound and outbound cargo handling, stack-layout adjustments, and space planning.

Warehouse Digitalization

Continuously accumulates inspection images, volume data, and spatial information, providing a foundation for digital warehouses and unmanned operations.

Recommended Validation Outcome: Complete an autonomous inspection of a designated warehouse area, output estimated cargo volumes, corresponding inspection images, and task records, and automatically generate a structured inspection report. 

02

Aerial Robotics Innovation and Creativity Challenge

The advancement of aerial robotics requires validation in more real-world scenarios. An innovative idea can only become a valuable product after development, testing, and application validation.

To enable more developers to take part, the 2026-2027 Aerial Robotics Innovation and Creativity Challenge is now officially open.

Open to universities, research institutes, technology companies, development teams, and aerial robotics enthusiasts, the Challenge provides an innovation platform integrating technical learning, project development, scenario validation, and project showcases.

Participating teams can use the FlyCore integrated control system for aerial robots to develop projects, progressing from solution design and system setup through scenario validation and competitive project demonstrations, and ultimately create their own innovative aerial robotics solutions.

The Challenge will also regularly organize aerial robotics development boot camps covering system setup, technology development, and hands-on project practice, providing participating teams with opportunities for learning and exchange as well as support for project development.

Ongoing Innovation Topics to Guide Exploration

To help participating teams move quickly into hands-on development, the Challenge will continue to release a series of innovation topics.

Each topic will combine an aerial robotics technology focus with a real-world application scenario, providing developers with project ideas and guidance.

Developers may conduct research based on the recommended topics or pursue further innovation according to their own interests.

Moving from a technical idea to an aerial robot that can truly perform tasks is exactly the innovation process the Challenge seeks to foster.

03

Build with FlyCore and Compete for a Million-Yuan Innovation Fund

The FlyCore integrated control system for aerial robots provides developers with an integrated development foundation spanning flight control, localization and navigation, and mission development.

FlyCore integrates core capabilities including flight control, mapping and localization, recognition and planning, and sensor integration. It establishes a complete closed loop spanning perception, localization, planning, and control, enabling developers to rapidly build aerial robotics platforms for algorithm validation and application exploration.

With FlyCore, participating teams can rapidly complete solution design, feature development, and scenario validation, turning their innovative ideas into working aerial robots that can truly fly and perform tasks. 

With FlyCore, teams can:

✓ Rapidly build an aerial robotics platform
✓ Validate algorithms and develop features
✓ Complete real-world scenario testing
✓ Present projects in the Challenge

Make ideas take flight and bring innovation into real-world applications.

This year's Challenge has established an annual dedicated incentive pool totaling RMB 1 million and will accept submissions on an ongoing basis for innovative aerial robotics projects with real-world application value and product potential. In addition to cash awards, the Challenge will provide outstanding projects with technical guidance, prototype refinement, engineering validation, supply-chain resources, and marketing support. High-performing projects may be selected for the product co-creation program and receive industry resources, startup incubation, or strategic investment support.