Bat-Inspired Miniature Drones: Revolutionizing Nighttime Search and Rescue Operations

Bats can contract and compress their muscles to listen only to specific echoes, a capability that inspires new drone technology.

Worcester:

Despite the fog machine, eerie lighting, and artificial bats, the robotics laboratory at Worcester Polytechnic Institute isn't celebrating Halloween. Instead, researchers are developing miniature drones capable of operating in search and rescue missions during darkness, smoke, or adverse weather conditions.

"Power lines are typically the first infrastructure to fail during earthquakes or tsunamis. Many disasters occur at night, and waiting until daylight to begin rescue operations isn't an option," explained Nitin Sanket, assistant professor of robotics engineering. "We began investigating nature to find organisms that could function effectively in these environments."

Sanket and his students identified bats as their inspiration, specifically focusing on their sophisticated echolocation abilities. With funding from the National Science Foundation, they're creating compact, affordable, and energy-efficient aerial robots that can operate in conditions where conventional drones cannot.

Recently, emergency responders in Pakistan deployed drones to locate individuals stranded on rooftops during extensive flooding. In August, a rescue team utilized a drone to discover a California hiker trapped behind a waterfall for two days. Additionally, in July, drones helped identify a secure path to three mine workers who remained trapped underground in Canada for over 60 hours.

While drones are increasingly utilized in search and rescue operations, Sanket and fellow researchers aim to advance beyond the manually operated individual units currently in use. According to Ryan Williams, an associate professor at Virginia Tech, a crucial development involves creating aerial robots that can operate in coordinated groups and independently determine search locations.

"The deployment of truly autonomous drones remains virtually nonexistent in current operations," Williams noted.

Williams addressed this challenge through a project that programmed drones to select search patterns in coordination with human rescuers. His team analyzed data from thousands of missing person cases to develop predictive models of behavior patterns for individuals lost in wilderness settings.

"We utilized these behavioral models to optimize drone positioning, focusing on areas with higher probabilities of locating missing persons," he explained.

At WPI, Sanket's research addresses additional limitations of existing drone technology, including size constraints and perceptual capabilities.

"Current robotic systems tend to be large, unwieldy, expensive, and unable to function across diverse scenarios," he stated.

In contrast, his drone is palm-sized, constructed primarily from affordable hobby-grade materials, and capable of operating in darkness. It employs a compact ultrasonic sensor, similar to those found in automatic faucets, to mimic bat behavior by emitting high-frequency sound pulses and utilizing the echoes to detect obstacles.

During a recent demonstration, a student remotely launched the drone in a well-lit environment and then again after reducing illumination to only a dim red glow. When approaching a transparent Plexiglas barrier, the drone consistently stopped and retreated, even in darkness with artificial fog and snow circulating through the air.

"Current search and rescue robots primarily operate during daylight hours," Sanket noted. "The challenge is that search and rescue operations are often tedious, hazardous, and difficult tasks frequently conducted in darkness."

Development encountered several challenges. The researchers discovered that the drone's propeller noise interfered with ultrasonic signals, necessitating 3D-printed shells to minimize interference. They also implemented artificial intelligence to help the drone filter and interpret sound data.

Despite these advances, significant development remains before matching bat capabilities. Bats can selectively contract and compress muscles to filter specific echoes and can detect objects as fine as human hair from several meters away.

"Bats possess remarkable abilities," Sanket acknowledged. "Our technology remains far from what nature has achieved. However, our objective is to eventually reach that level and develop systems suitable for deployment in real-world environments."