A researcher at Worcester Polytechnic Institute (WPI) is designing small flying robots modeled after bats to support search and rescue teams in places where smoke, dust, rough terrain, or harsh weather make human entry difficult. The project aims to create robots that can move through tight spaces, detect obstacles at close range, and operate with minimal power demands by copying how bats sense their surroundings.
Professor Nitin J. Sanket and his team have built palm-sized robots that rely on ultrasound, mirroring how bats emit sound and interpret returning echoes. The robots use AI software to remove noise from the ultrasound signals so they can detect obstacles within a two-meter radius. Sanket told TechCrunch that current rescue operations often require people to walk through dangerous environments with flashlights, and his team sees drones as a way to reduce risk because they can move quickly and cover wide areas.
Sanket’s interest in aerial robots began during his PhD program, when his adviser challenged him to build the smallest robot possible. That assignment pushed him toward studying how insects and birds perform complex flight tasks with limited computing power and relatively simple sensory systems. He noted that these animals manage navigation despite small brains and modest visual systems, prompting his shift toward biology-inspired engineering and eventually shaping his doctoral research.
One of his early experiments involved a robotic beehive made of miniature drones designed to pollinate flowers. Although the system worked at a prototype level, Sanket later assessed that the idea would take too long to become practical and began looking for applications where bio-inspired robotics could have impact sooner. Search and rescue presented clearer opportunities due to the need for small, low-cost, low-power devices that can navigate hazardous locations.
A major challenge came from balancing sensor requirements with size and energy limits. The team adopted ultrasound components similar to those found in automatic faucets because these sensors use very little power. However, the noise produced by the robots’ propellers interfered with the sensor readings and blocked the devices from accurately detecting obstacles.
The solution again came from studying bats. Sanket explained that bats have tissues in their nose, ears, and mouth that shift thickness and density to adjust how they emit and hear sound. His team replicated this concept by creating a 3D-printed structure positioned in front of the robot to alter the path and shape of the sound waves. Functionally, this allowed the robot to distinguish its own propeller noise from environmental reflections, improving detection accuracy.
With the system now functioning, the team’s next priority is increasing flight speed. Sanket said that researchers often try to emulate human cognitive abilities in machines but overlook how effectively animals perform navigation with far smaller physical and computational resources. He suggested that studying insects and birds can offer insights that engineers might miss when focusing solely on human-like behavior.
Featured image credits: Roboflow Universe
For more stories like it, click the +Follow button at the top of this page to follow us.