A plan to develop tiny, autonomous flying robots that can navigate in total darkness, similar to how bats use sound waves to map their terrain, has been awarded funding from the National Science Foundation (NSF).
Professor Nitin Sanket, a robotics engineering scientist from Worcester Polytechnic Institute (WPI) and the project lead behind the ambitious program, says that the ability of these robots to use ultrasound-based echolocation instead of line-of-sight navigation makes them ideal support tools for search and rescue operations in dark, dusty, and smoky disaster areas, as well as for remote monitoring in other challenging environments where cameras or other light-based sensing platforms fail.
“We believe this is the first step for widespread deployment of drones for search and rescue in harsh and critical conditions such as wildfires, earthquake zones, and volcanoes,” explained Professor Sanket in an email to The Debrief.
Inspired by Bats
Sanket told The Debrief that the recently awarded grant from the National Science Foundation’s “highly competitive” Foundational Research in Robotics (FRR) program will allow him and his team to “draw inspiration” the behavior of bats, who use echolocation to create a mental sound map of the local terrain even in limited light, fog, or other circumstances where navigation by sight is impossible.
“We will not explicitly study bats in our work but rather draw inspiration from large bodies of existing and future work,” the researcher explained, adding that his team’s goal is to “mathematically model” the findings in real bats discovered by other ecologists. These models will inform the material and software design of their final robot prototypes.
In-House Production of Next-Gen Drones
According to the terms of the award, Sanket’s team will spend the next three years building and testing tiny aerial robots that are smaller than 100 millimeters and weigh under 100 grams. Sanket told The Debrief that his team will not use off-the-shelf drones but instead will build all their robots “in-house.” They will also explore different metamaterials that minimize noise interference from the drone’s propellers. Minimizing the noise from the drone’s propellers increases the ability of the ultrasensitive ultrasound sensors to detect reflected sound waves to locate and avoid otherwise obscured obstacles.
“The mechanical structure is designed and manufactured from scratch with some off-the-shelf avionics components like the flight controller, Electronic Speed Controllers (ESCs), and motors,” Professor Sanket explained. “The ultrasound sensors are manufactured by TDK electronics who we collaborate with to build mathematical models and 3D printed structures for noise reduction.”
Perhaps the most challenging part of the work is building AI software that utilizes the ultrasound data captured by the sensors to navigate in the dark or other obscured-sight environments. Sanket told The Debrief that the combination of the drone’s hardware design and this built-from-scratch software stack “couples mechanical noise reduction with AI models for obstacle detection in the presence of extreme noise (from propellers and environment)” to operate in challenging and harsh conditions. This work will also draw inspiration from bats’ abilities and the physiognomy they have evolved to enhance those abilities.
“We will build AI models for how bats process ultrasound, along with mechanical design inspired by bats’ noseleaves and mouth for noise reduction and targeted ultrasound echolocation,” he explained.
Efficiency and Ultrasound Sensors
Along with the autonomous navigational benefits of using ultrasound instead of cameras or LIDAR, Professor Sanket pointed to the technology’s significant power savings. Reducing power usage can dramatically extend missions and free up energy for additional tools and sensors on future drone models.
“Our work presents a revolutionary new method for using extremely noisy ultrasound sensors using only 1 milliwatt of power,” Professor Sanket told The Debrief, noting that this value is close to one thousand times less than cameras and LIDAR systems.
When asked if the drones’ flight methodology and characteristics would also be based on bats, the researcher said the first prototype would be built on custom-built palm-sized quadrotors. However, he added that in the project’s second year, they will “explore other modalities for propulsion” such as flapping wings to mimic birds and insects.
“We believe this should generate less ultrasound noise compared to propellers,” Sanket told The Debrief.
Once their three-year term is completed, the professor said his team’s research will be made “open source for future development.” When discussing potential applications of tiny, autonomous drones that can navigate in total darkness, the professor noted that drones with these capabilities could enable “rapid deployment of robots in challenging environments,” including disaster zones and smoke-filled areas. This technology could also lead to the development of autonomous drones for environmental protection and monitoring potential disaster areas.
“Our research is the first to showcase working in visually degraded conditions such as smoke, dust, darkness, and snow on a palm-sized robot,” Sanket told The Debrief. “Furthermore, this brings a fresh perspective on autonomy by utilizing ultrasound, which remains heavily under-explored in flying robots due to their inherent propeller noise.”
Beyond powering tiny robots that can fly in total darkness, the WPI professor said his team’s work could enable advances in sound-based navigation for self-driving cars, coral reef monitoring, and the exploration of dangerous or challenging areas, such as volcanoes. However, he admitted that his broader vision involves a world where tiny, flying autonomous drones aren’t simply machines but instead operate as robot “partners” to improve several areas of human life.
“I envision a future where we are surrounded by small robots doing various tasks such as pollinating flowers, searching for survivors, carrying payloads, entertaining us, and even as our personal pets/companions,” the professor explains in his WPI bio.
The project “Sound Navigation: Enabling Tiny Robots to Find Their Way Through Smoke, Dust, and Darkness” is funded by a grant from the National Science Foundation’s Foundational Research in Robotics (FRR) program.
Christopher Plain is a Science Fiction and Fantasy novelist and Head Science Writer at The Debrief. Follow and connect with him on X, learn about his books at plainfiction.com, or email him directly at christopher@thedebrief.org.