Remote Sensing at a Crossroads: Growth and Critical Reflection

Remote sensing as a discipline is currently experiencing what appears to be a period of rapid expansion. More programs are being established across universities, more institutes are opening, and the field seems to be flourishing. Yet, behind this appearance of prosperity, there is a need for critical reflection.

Much of today’s research remains confined to traditional, paper-oriented agendas. Projects are often locked in cycles of repetition and incremental improvements, frequently measured by “state-of-the-art” benchmarks on fixed datasets or exaggerated claims of novelty. While such work can still appear in top journals, it rarely leads to groundbreaking insights or meaningful applications. Too often, the field risks falling into a loop of academic self-entertainment. The irony is striking: in practice, even government agencies often rely on manual labeling rather than automatic algorithms for remote sensing–based recognition. This gap highlights a regrettable shortcoming.

Of course, these tendencies are shaped by broader realities. Few scholars can remain completely independent of institutional and systemic pressures. Still, it is important to acknowledge that remote sensing itself is highly valuable. Observing the Earth’s surface from space over large areas and long time spans is both powerful and inspiring, with fundamental importance for defense, environment, ecology, urban development, and beyond. The problem lies not in the discipline’s potential, but in how research has become skewed—overly focused on narrow, paper-driven directions rather than real-world problems.

In my view, remote sensing should give greater emphasis to what might be called “broad remote sensing and space engineering”: satellites, sensors, imaging systems, orbital platforms, and space-based perception. The focus should be on acquiring more timely, higher-quality data and solving pressing, practical problems. Once reliable data exist, a variety of methods can be developed to apply it. The obsession with post-hoc algorithmic competition often misses this point.

Admittedly, such opportunities are not always easy to realize in the current environment. But promising intersections exist. Remote sensing can find powerful applications in emerging domains such as autonomous driving and spatial perception, high-definition map construction and updating, or the low-altitude economy supported by drones and aerial logistics. More broadly, there is space for scholars to move beyond traditional boundaries, engaging with computer science, automation, navigation, control, and planning. Building practical and deployable algorithms from the perspective of real applications will allow remote sensing to serve daily life, not remain confined to the laboratory.

I am increasingly convinced that if remote sensing is to avoid marginalization, it must leave its academic comfort zone, engage more actively with industry, and embrace interdisciplinary collaboration. Only then can the field regain vitality and direction.

The road ahead may be difficult, but I believe the future is promising. Paths will open, and with collective effort, new breakthroughs will emerge. I hope more scholars will join in exploring these possibilities.