Thanks to their highly sensitive whiskers, which enable them to detect objects within reach before they get too close, rats and mice can hardly relate to the pain of a stubbed toe, despite having eight more toes than humans. This sensory organ lets rats scurry through buildings, sewers, burrows, and countless other environments with ease, regardless of visibility. Beyond touch, whiskers can also detect vibrations and even subtle air currents.
Inspired by this tactile sensing, researchers at Delft University of Technology in the Netherlands have developed what they describe as a lightweight, whisker-based tactile sensor that gives drones the ability to navigate and explore their surroundings through gentle touch. Made from nitinol wire, an ultralight shape-memory nickel-titanium alloy, the whiskers let the drones sense their surroundings through touch rather than sight, allowing them to operate in darkness and low-visibility conditions.
Now, today's drones don't randomly crash into objects. Modern autonomous drones are highly sophisticated pieces of equipment with advanced navigation and object-avoidance systems that ensure they almost never run into obstacles. The issue, however, is with tiny autonomous drones. Below 100 grams, it becomes much harder to balance weight constraints with the performance these systems offer. Cameras, LiDAR, and other conventional navigation sensors are usually too heavy to integrate into such aircraft.
To address this challenge, the researchers created lightweight, pressure-sensitive artificial whiskers, each weighing only 3.2 grams and measuring 200 mm long. They integrated two whiskers at the front of the drone. At the base of each whisker filament are three miniature barometric-pressure sensors connected to an onboard microcontroller. When the drone flies near an object and a whisker grazes or bends against a surface, the pressure changes at the base are measured in under two milliseconds, enabling the drone to calculate contact location and depth with millimeter precision.
“Here, we aim to equip drones with rich tactile sensing – not for manipulation in the air, but for a novel concept of tactile navigation: using touch to explore and fly through the unknown. But this comes with a challenge: for tactile sensing to work on drones, it needs to be lightweight, low-latency, and low-power. Inspired by nature, we found the answer in whiskers,” explains Salua Hamaza, Associate Professor of Aerial Physical Interaction & Embodied Intelligence in Aerial Robots at Delft University of Technology.
Now, you may be wondering how a highly sensitive pressure sensor attached to a pressure-generating component like a drone that vibrates and produces noisy airflow would work. The researchers thought of this as well. They developed a lightweight, real-time processing pipeline that uses advanced onboard software to distinguish meaningful touch information from the turbulence generated by the drone's own propellers, extracting reliable tactile information only during flight. Impressively, the system does this with a mere 34 kilobytes of data.
“We wanted to show that touch does not have to come at the cost of size or computational power. Our entire tactile perception pipeline runs onboard using just 34 kilobytes of memory, allowing a tiny drone to sense and respond to its environment in real time,” said Chaoxiang Ye, a participant on the project.
In experiments, the whisker-equipped drone not only followed surface contours in complete darkness and explored enclosed spaces through wall-following behavior, but also built tactile maps of unfamiliar environments and located exits without relying on vision-based sensors. These abilities are very useful in search and rescue scenarios where darkness, smoke, dust, and other particulate matter can limit vision-based navigation.
An important caveat here is that the whiskers need to actually touch an object to know that it's there. Considering that the whiskers are only 200 mm long, this means that the drones cannot go zipping through unmapped terrain at high speeds. Its momentum will ram it straight into the obstacle before the whiskers detect it's even there.
Furthermore, the whiskers are in front, preventing the drone from detecting sideways obstacles like random jagged branches in a forest. The system is purpose-built for structured, low-visibility tunnels, and excels in environments with continuous, relatively flat surfaces, where the drone can press its whiskers against a wall and follow it like a vision-impaired person using a cane.
The study was published in the journal Nature Communications.
Source: Delft University of Technology
