Engineering students at the University of Cincinnati are using sound to help NASA detect if a drone's motors or propellers are damaged to prevent a crash.
UC College of Engineering and Applied Science doctoral student Prissha Chawla and her research partners recorded the various sounds that drones make when their propellers get chipped or cracked or their motor becomes damaged or begins to fail.
They converted the recordings into valuable data that computers at a ground station can use to alert operators that a drone is not airworthy before it takes off.
"The project started in 2024 with interest by NASA in detecting problems with motors before a drone took off," Chawla said. "More and more, drones are being used for public safety. You don't want them to crash."
National media outlets including Yahoo! News highlighted the students' projects.
It was one of several projects in Professor Manish Kumar's lab at Digital Futures that UC students presented at the Modelling, Estimation and Control Conference Oct. 25 in Phoenix, Ariz.
UC Professor Manish Kumar works with students in the high-bay lab at Digital Futures. Photo/Michael Miller
Drones navigate warehouse using digital twin
For another project, UC students are developing an easy-to-use system to deploy autonomous drones in warehouses to take inventory and alert workers about leaks from containers and other urgent issues.
Doctoral students Deepak Antony David and Luke Busse are working with Valvoline to use drones to log inventory and detect leaks that could damage entire pallets.
Instead of navigating by satellite, the indoor drones use a tool more commonly associated with geology: light detection and ranging. The drone emits pulses of laser light for obstacle avoidance within centimeters of accuracy.
With high fidelity, it looks like a video game.
Luke Busse, UC College of Engineering and Applied Science student
But what makes the system unique is the drones are dispatched using a photorealistic model of the warehouse floor called a digital twin. Workers can explore the virtual spaces of the warehouse interior depicted in high-definition video. At the click of a mouse, they can send an autonomous drone to a particular aisle or section of the warehouse to perform tasks such as finding items or looking for damaged pallets.
The students presented their proof of concept at the national conference, sponsored in part by UC's College of Engineering and Applied Science.
"The big problem with inventory is inaccuracy," Busse said. "We found that 65% of warehouses in North America still have manual inventory. People picking products, categorizing things wrong or putting them in the wrong spot. It takes time to fix."
UC doctoral students Luke Busse, pictured, and research partner Deepak Antony David presented their proof of concept at the Modeling, Estimation and Control Conference this month in Phoenix, Ariz. Photo/Michael Miller
Dispatching drones using virtual model
The autonomous drones can navigate the tight indoor confines of warehouse aisles. And they can reach the highest shelves quickly and easily.
"A drone doesn't have trouble with the third dimension," Busse said.
UC's drone research: Frequently Asked Questions
What drone projects are UC students working on? right arrow down arrow
Mechanical engineering students at the University of Cincinnati are using sound to help NASA detect if a drone's motors or propellers are damaged to prevent a crash. They are converting the sounds of damaged propellers and motors into datat the computers can use to warn operators that a drone might not be airworthy.
UC is also working with Valvoline to develop autonomous drones that use light detection and ranging to navigate the tight confines of a warehouse. The project uses a virtual model of the warehouse called a digital twin to help workers dispatch the drone to carry out tasks.
And UC is developing a telehealth project to use drones to reach patients who might not be able to make an in-person visit to a doctor.
Can drones tell if they're damaged? right arrow down arrow
Yes, some drones can tell if they are damaged. UC researchers recorded examples of chipped, cracked and otherwise damaged propellers and common motor issues that can lead to seizing and converted the sounds to data that a computer can use to warn of potential issues before a drone takes off. Eventually, the students envision an autonomous drone that can use sound alone to detect damage to its own propellers and other mechanical systems in flight so it can return to base before sustaining a catastrophic failure.
Are drones used in warehouses? right arrow down arrow
Yes, drones work in many warehouses. UC's College of Engineering and Applied Science in Cincinnati, Ohio, is working with Vaolvoline to develop autonomous drones that can conduct inventory or spot issues such as damaged products.
Can I work with drones in college? right arrow down arrow
Yes, the University of Cincinnati allows students to work on a variety of drone research projects or develop their own using facilities such as the multistory high-bay lab at Digital Futures, where students conduct flight testing.
UC doctoral student Prissha Chawla and her research partners are using sound to diagnose damage to drones before takeoff. Photo/Michael Miller
Sound provides telltale clues of possible drone damage
As drones increasingly find new uses in public airspace, UC's Chawla said it is important to engineer better ways to prevent accidents.
Eventually, the students envision an autonomous drone that can use sound alone to detect damage to its own propellers and other mechanical systems in flight so it can return to base before sustaining a catastrophic failure.
Besides chips or cracks in propellers, Chawla said grit and other debris can get into motor bearings, causing it to seize.
"We're focusing on pre-flight tests that NASA already runs," she said.
UC's Busse said the damage-detection system could be a feature of fully autonomous "drone-in-a-box" systems. Drones deploy from a weatherproof docking station that handles launch, landing, charging, data upload and mission scheduling without an on-site operator.
"In that scenario, it would be great. You could have sensors in the box that provide preflight checks," Busse said.
Featured image at top: UC doctoral student Prissha Chawla inspects a drone in Professor Manish Kumar's lab at Digital Futures. Photo/Michael Miller
