Drones now fly in precise patterns without GPS, thanks to a new control approach based on what they can see with their cameras.
Researchers at IIT Bombay have introduced a bearing-only control scheme that uses onboard camera data to determine relative positions and keep a formation, without GPS, inter-drone chatter, or a centralized controller.
The work focuses on Vertical Take-Off and Landing (VTOL) aircraft, which can hover and operate in tight spaces, expanding the reach of multisensor surveillance and inspection missions.
Recent Trends
- GPS-denied drone operations rising in both military and civilian sectors
- Vision-based formation control attracting researchers and startups
- VTOL drones expanding use in confined spaces like indoor inspection
In bearing-only control, each drone observes its neighbors via bearing measurements—the direction to each peer—rather than distances. The team has proven the approach mathematically, ensuring convergence to the desired formation even when starting from imperfect positions.
They analyzed two operational scenarios: one with constant velocity using bearing data; another with time-varying formation and velocity, where each drone uses its own velocity along with bearing data. The method also handles arbitrary time-varying configurations, enabling drones to pass through narrow gaps, reconfigure into linear layouts, and adapt to mission changes.
According to The Hindu, the researchers, led by Professor Dwaipayan Mukherjee and research scholar Chinmay Garanayak, plan to test the scheme experimentally with a drone swarm and aim to add theoretical collision avoidance guarantees in future work.
Why this matters: GPS-denied environments, electronic warfare contexts, and stealth operations can benefit from sensor-local control. Using bearing rather than distance measurements reduces hardware weight and battery draw, while still enabling robust formation maintenance, which can be a game changer for search and surveillance in constrained spaces. For industry readers, the shift toward decentralised, sensor-first autonomy is already reshaping how we design swarms for real-world tasks.
What makes bearing-only control different
Bearing-only control relies on the angles to neighbours rather than precise distances. Think of it like following compass directions rather than map scales. This makes the system more robust to noisy range sensors and reduces sensor payload, which helps extend flight times on small VTOL platforms.
Implications for industry and defense
Decentralised, camera-based formation control aligns with growing demand for autonomy in GPS-denied zones. In practice, it could improve covert surveillance operations, search missions, or complex logistics corridors where signals are unreliable or jammed. VTOL configurations, which naturally lift off in tight spaces, stand to gain the most as swarms become easier to deploy indoors or in urban canyons.
Next steps and timelines
The IIT Bombay team plans to move from theory to field tests with real drone swarms. They also want to prove collision avoidance with rigorous guarantees, addressing a common gap in practical swarms where safety depends on ad hoc rules. For practitioners, this work suggests a design path: prioritize local bearing data and robust formation plans that do not rely on central coordinators.
Conclusion
The push toward GPS-free autonomy is accelerating. This bearing-only approach adds a rigorous, camera-based option to the menu, broadening where and how drone swarms can operate. If validated in real-world flights, it could redefine how operators think about surveillance, inspection, and logistics in GPS-challenged environments.
