Drones Radar

MIT Drone, MiFly, Indoor Navigation Using Radar

Credit: Signal Kinetics

Image Credit: Signal Kinetics

Navigating drones indoors has always been a challenge. With GPS signals unavailable, computer vision struggling in low-light or featureless spaces, and LiDAR systems being both expensive and power-hungry, the need for a new approach was clear. MIT researchers have taken a bold step forward by leveraging millimeter-wave (mmWave) technology—a form of radar—to overcome these obstacles. For a quick reminder on how radar differs from LiDAR, visit What Is The Difference Between Lidar and Radar?.

The Technology Behind MiFly

The MiFly system uses mmWave RF signals similar to those found in 5G networks and traditional radar systems. Instead of depending on external infrastructure, MiFly uses a single low-power backscatter tag—essentially a sticker that reflects RF waves—placed in the environment. This innovative approach enables the drone to self-localize with remarkable accuracy. The obvious limitation here being that MiFly can only operate in environments with backscatter tags installed. For structured environments that are anticipating the use of such technology, though, this could be a significant step forward.

Equipped with two off-the-shelf radars—one horizontally polarized and the other vertically—the drone emits RF waves that reflect off the tag. To ensure that the tag’s signal stands out from environmental noise, the researchers implemented a dual-modulation technique. The tag adds a distinct frequency to the reflected signals, allowing the drone to filter out irrelevant data. When this information is combined with readings from the drone’s inertial measurement unit (IMU), which tracks acceleration and orientation, MiFly achieves a complete six-degree-of-freedom pose estimation (pitch, yaw, roll, and spatial coordinates) within milliseconds.

Performance and Testing

Tests conducted in various indoor settings—including MIT laboratories and tunnels—demonstrated that MiFly can achieve localization accuracy within 2.76 inches (7 centimeters) and maintain effective performance up to 19.7 feet (6 meters) from the tag. These impressive results highlight the potential of mmWave-based positioning to operate reliably even in challenging environments.

Industry Implications

The breakthrough achieved by MiFly has far-reaching implications. From warehouse operations and industrial inspections to search-and-rescue missions and autonomous indoor navigation, this technology offers a simple, cost-effective solution for precise indoor positioning. As drones continue to evolve into fully autonomous machines, RF-based localization could become a key enabler for more versatile, reliable, and efficient operations.

For further details on the technology and its development, visit the MIT Media Lab project page on Indoor Drones That Can Fly in the Dark.

Could RF-based positioning be the key to unlocking widespread indoor drone operations? Let’s discuss in the comments below.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.