How long does it take to turn a table full of loose components into a drone that can lift off and hover steadily?

At the FlyCore Aerial Robot System Design & Integration Bootcamp, the answer was four days.

Power on. Arm the system. Push the throttle. The moment the rotors began to spin, four days of assembly, wiring, troubleshooting, and tuning finally paid off: the drone built by the participants themselves was flying steadily.

From August 25 to 28, the FlyCore Aerial Robot System Design & Integration Bootcamp was held in Chengdu. Participants from different backgrounds formed teams and started with a complete set of disassembled B320 drone components, working step by step through full-system assembly, system debugging, sensor calibration, and indoor position-hold flight verification.

By the end of the bootcamp, every team had completed two deliverables: a B320 that had passed basic flight verification, and a product definition proposal tailored to their own industry or application direction.

Over the course of four days, the program progressed from system-level understanding to full-platform assembly, then to debugging and verification, and finally to product solution development. Everything learned during the program was continuously tested through hands-on work with real aircraft and repeated refinement of product concepts.

01

What Did Participants Accomplish in Four Days?

DAY 1: Understand the Complete System First

The first day focused on answering one fundamental question: Why is the aircraft designed this way?

The course began with the OODA control framework, helping participants understand the complete operational loop of an aerial robot—from observing and sensing the environment, to making decisions, and finally executing actions. Flight controllers, payload computers, sensors, communication systems, and propulsion systems were then examined as parts of one integrated architecture.

Using the SU17 and U550 test platforms as examples, system integration engineers analyzed typical engineering considerations such as:

• weight and power consumption,
• equipment placement and center of gravity,
• sensor field of view and occlusion,
• power distribution,
• and communication topology.

Participants then began drafting their own product definition documents. They were asked to answer several key questions:

• What is the intended application scenario?
• How will the mission be carried out?
• How should the key performance requirements be defined?
• How will the final system be verified?

These questions continued throughout the following three days and also formed the core of the final solution review.

Product definition directly affects subsequent component selection. Payload weight influences propulsion requirements and endurance; mission distance affects the communication solution; the operating environment determines sensor configuration; and the verification method determines the testing plan.

Only when the objective is clearly defined can assembly and debugging be carried out with clear engineering criteria.

---

DAY 2: Start with Loose Components and Build a B320 by Hand

On the second day, the B320 was placed on the workbench as a collection of individual components:

• airframe,
• propulsion system,
• power distribution board,
• ESCs,
• FlyCore upper and lower boards,
• LQ3 video and data transmission module,
• cables,
• connectors,
• and various mechanical components.

Each team worked through the assembly process step by step, including connecting the power distribution board and ESCs, assembling the upper and lower boards, installing the video and data transmission module, and organizing the complete wiring harness.

In this way, the system architecture learned on Day 1 was translated into every physical connector and component on the aircraft.

Full-system assembly requires simultaneous consideration of power distribution, communication, and maintainability.

Whether cable routing is clear, connectors are securely installed, and the power topology is correct can all directly affect subsequent system debugging.

Before powering on the aircraft, each team used a multimeter to check for short circuits. Once the power and communication systems were confirmed to be functioning correctly, a complete B320 aerial robot had taken shape on the workbench.

---

DAY 3: Configuration, Calibration, and Position-Hold Flight

On the third day, the completed aircraft entered the configuration and flight verification stage.

Each team completed:

• LQ3 video and data link configuration,
• flight controller parameter setup,
• sensor calibration,
• and indoor position-hold flight verification.

Following the workflow of PX4 multirotor tuning, the flight-control development engineer explained how PID parameters affect aircraft behavior and how flight logs can be used to identify vibration and control oscillation problems.

The course organized the debugging process into a clear sequence:

Phenomenon → Log Data → Root Cause → Adjustment

When an abnormality occurs, participants were encouraged to first look for clues in the data, then inspect the structure, sensors, connections, or control parameters.

This approach helps reduce repeated trial-and-error caused by blindly changing parameters.

A vision development engineer also delivered a dedicated session on multi-sensor calibration, introducing the workflows of two toolchains, Kalibr and FAST-Calib, and guiding participants through a hands-on camera calibration exercise.

Participants needed to pay attention to every stage of the process:

• collecting data,
• running the calibration,
• reviewing the results,
• and evaluating whether the calibration output was reasonable.

This allowed abstract coordinate relationships to be translated into actual data and procedures that could be inspected and verified.

The stable operation of a multi-sensor system depends on the accuracy of both spatial coordinate relationships and timing relationships.

Calibration results continue to affect localization, perception, and mission execution. Therefore, checking calibration data and evaluating the results after calibration are also critical parts of full-system debugging.

---

DAY 4: Turn Hands-On Experience into a Product Solution

On the fourth day, the course returned to the industry application directions first proposed on Day 1.

Drawing on three days of hands-on experience with real aircraft, each team continued refining:

• airframe and propulsion selection,
• structural layout,
• electrical and communication architecture,
• software architecture,
• and verification plans.

The goal was to transform an application idea into a relatively complete aerial robot product solution.

In the afternoon, each team presented its proposal in turn and responded to questions and reviews from the instructors and other participants. An outstanding solution team was also selected from the bootcamp.

By the end of the program, every team had:

• a B320 that had completed basic flight verification, and
• a product definition proposal that could continue to be refined and tested.

Whether they move on to participate in the challenge competition or continue developing their own industry projects, the participants now have a much more concrete starting point.

---

02

The Engineering Decisions Behind a Real Aircraft

When an aerial robot enters a real-world application, many engineering issues emerge at the same time.

Adding cameras, computing platforms, or other payloads changes the aircraft's:

• total weight,
• power consumption,
• and endurance.

Changing equipment positions affects the center of gravity.

Changing the camera mounting angle affects the field of view.

Communication links and power architecture directly influence overall system stability.

All of these factors interact with one another. Looking at any single specification in isolation is rarely enough to explain the performance of the complete aircraft.

The bootcamp brought all of these issues together on the same real aircraft.

From checking for short circuits before power-on, to analyzing flight logs after testing, performing camera calibration, and reviewing product solutions, participants were required to observe what was happening, read the data, inspect the physical connections, and then return to the original product objectives to make engineering trade-offs.

Once engineering decisions are supported by evidence, subsequent adjustments become much more focused and effective.

---

03

After the Bootcamp, the Challenge Continues

The 2026–2027 FlyCore Innovation Challenge is currently accepting project submissions.

The competition is open to:

• UAV technology developers,
• corporate R&D professionals,
• university innovation teams,
• and industry partners.

It focuses on identifying aerial robot solutions that address real-world needs and demonstrate both application value and product potential.

High-risk environments, hard-to-reach locations, repetitive operations, and other specialized scenarios may all provide opportunities for new innovations.

The FlyCore Integrated Control System combines foundational capabilities including:

• flight control,
• mapping and localization,
• recognition and planning,
• and intelligent computing.

It provides teams with a technical foundation for further development.

Participants can therefore devote more of their effort to:

• AI algorithms,
• mission logic,
• industry applications,
• and user requirements,

while progressively refining their solutions through prototype verification.

Who Can Participate?

Universities, research institutes, technology companies, and UAV developers are all eligible to register.

Participating teams are required to carry out independent development and original design based on the FlyCore Integrated Control System.

What Kind of Projects Can Be Submitted?

Projects may focus on applications such as:

• heavy-load operations,
• inspection,
• tunnels,
• exploration,
• warehousing,
• utility corridors,
• special operations,

or other scenarios familiar to the development team.

Each project should clearly define:

• the target operation,
• environmental constraints,
• and the core problem that needs to be solved.

Teams should then use prototypes, test data, or on-site demonstrations to prove that their solution can successfully complete the intended mission.

What Support Does the Competition Provide?

The current FlyCore Innovation Challenge has established an annual incentive fund totaling RMB 1 million.

In addition to cash awards, outstanding projects may receive support in areas including:

• technical guidance,
• prototype refinement,
• engineering verification,
• supply-chain resources,
• and marketing and promotion.

High-performing teams may also have the opportunity to enter a product co-development program and gain access to:

• industry resources,
• startup incubation,
• or strategic investment support.

The goal of the competition is to extend a one-time project showcase into sustained product development and long-term industry collaboration.