In this edition, we speak with Wu Guanghui, an application software engineer who has spent years in the automotive sector. For the past two years, he has walked into the same office, opened MATLAB, built models, generated code, and debugged logic—routines that look identical to his previous work. But he knows the product he is building is no longer simply a ‘car’.

He’s building a flying car.

Stowing an Aircraft in the Trunk

"The product we’re developing allows the aircraft to be stowed inside a car. When flight is needed, the module detaches and conducts low-altitude operations." It sounds like science fiction, yet in his vocabulary it is merely a project brief. As an application-layer software engineer, his task is to build control models in MATLAB/Simulink, auto-generate code, and deploy it to the flight-control system.

This is low-altitude flight.

Efficiency Practices in Application Development

When developing I/O models for the aircraft, the signal interface count typically runs into the thousands. Manually building models from scratch involves a large workload and high repetition, as each interface must be defined, connected, and checked. Any errors can only be detected during the later verification phase, resulting in significant rework costs.

His solution: MATLAB scripts that auto-generate the required interface models. "Generating models automatically through scripts can significantly enhance efficiency." This approach frees engineers from repetitive tasks, redirecting focus toward control logic and system architecture.

Sound Architecture Is the Prerequisite for Efficiency

Automation solves repetition, but in his view, it is not the decisive factor. What truly determines project trajectory is the soundness of the model architecture.

"The architecture must be well-structured," he says, "because all subsequent feature enhancements, maintenance, and redevelopment rests on that foundation. If the architecture is flawed, you may have to scrap and start over."

For low-altitude aircraft, one principle is non-negotiable: signal inputs, signal outputs, and every intermediate logic stage must be explicit and traceable. Compared with automotive systems, aircraft development tolerates far less error. Iterating on unstable architecture only compounds problems. This isn’t a matter of style. It’s an engineering baseline.

The Technical Opportunity Presented by Low-Altitude Aircraft

The sector is currently riding a wave of policy momentum. In 2026, the low-altitude economy was designated as a priority emerging-industry direction for the 15th Five-Year Plan period, with multiple cities rolling out dedicated policies and accelerating the build-out of airspace management frameworks and infrastructure. Wu confirms the trend: "This is a sector the state is actively promoting, and many cities are driving it forward."

For application software engineers, low-altitude aircraft present a natural extension of existing skill sets. Demands for flight-control algorithms, sensor fusion, automatic obstacle avoidance, and path planning overlap heavily with intelligent automotive systems, yet impose stricter requirements on real-time performance, safety integrity, and redundancy design.

The core logic remains unchanged, but the technical complexity has stepped up. In the past, an automotive application-layer engineer’s career path oscillated between legacy OEMs and new EV players. The emergence of low-altitude aircraft effectively branches that same skill tree. Existing competencies in modelling, code generation, and control logic development transfer directly, while presenting higher-order technical challenges.

"For engineers, low-altitude flight is a strong growth vector." This is not a passing comment on market trends. It’s an engineer’s affirmation of his own capability boundaries: the logic that holds on the ground still holds in the air, only the tolerances are tighter.