A team of Binghamton University researchers has developed a sensor system aimed at helping medical professionals more efficiently monitor how many people are coughing – and potentially spreading a disease – within a given area.
The prototype, dubbed " CoughNet ", was developed by Binghamton University Assistant Professor Dali Ismail , from the University's Thomas J. Watson College of Engineering and Applied Science , as well as PhD students Amir Esmaeili and Maryam Fazli. The prototype was presented at the 2026 Institute of Electrical and Electronics Engineers/Association for Computing Machinery Conference on Connected Health: Applications, Systems, and Engineering Technologies (CHASE) in August 2026.
CoughNet is a series of “listening stations” that detect coughs, determine if the sound was indeed a cough, and then check whether the cough came from a person who was already recorded or from a new individual.
To accomplish this, three Raspberry Pis — small, inexpensive single-board computers with built-in microphones — are placed in a room. When someone coughs in the room, all three hear the sound, but the one closest to the person who coughs hears it the loudest.{!--[aside]1[/aside]--}
The computer then captures a short recording of the cough and runs the sound file through an artificial-intelligence model that determines whether the sound is an authentic cough. To decide who the cough came from, the system analyzes the recording from all three microphones, examining how strong the cough was in each recording and how long the sound took to reach it.
“The microphone closer to the source acts as a reference microphone, and we can do correlation to determine if this cough is from the same exact person, because the same cough will be heard by the two other microphones, which are a little bit far away,” Ismail said.
After a series of coughs are recorded, the system is able to detect a pattern. If multiple coughs feature a similar loudness and time to reach the microphone, those coughs likely came from the same person.
The COVID-19 pandemic held up a magnifying glass to respiratory illnesses and how they spread. Health professionals and researchers started paying greater attention to the risk factors of being in the vicinity of sick individuals. Ismail saw this as the spark to develop the system.
“Most of the work on public health and disease monitoring has been very machine learning‑heavy and processing‑intensive,” said Ismail, a faculty member at the Thomas J. Watson College of Engineering and Applied Science’s School of Computing. “We wanted to look at cough detection and identifying who is coughing as a lightweight application that could eventually run on practical devices.”
Efficient with an eye on privacy
The system is not only effective but efficient. Raspberry Pi computers are extremely affordable; many models can be purchased for less than $50. On top of that, these computers are extremely energy-efficient, and that comes down to how the computers transmit the sound files.
The devices use low-power, long-range (LoRa) wireless technology, which covers far more indoor space than a single Wi-Fi access point while using so little power that sensors can run for years on batteries.
Because LoRa can only handle small messages, the units are selective about what they send. If a cough recording is clean, they process and group it locally. If it's noisy or unclear, they offload a small clip to a nearby, more powerful computer. In short, CoughNet processes what it can and offloads only what it must.
“It's basically a combination of networking and computation. Both are optimized to efficiently work on those tiny devices, and we also went with LoRa because it's extremely low-energy compared to Wi-Fi,” Ismail said. “A typical lower sensor can run on AA batteries for 10 years or so.”
With medical settings being likely use for CoughNet, it raises an obvious privacy question. Ismail notes that CoughNet only counts the number of unique coughs in the room and does not identify the people making them.
Unlike other public health sensing systems that rely on thermal cameras or visual sensors, CoughNet reacts only to short bursts of sound resembling a cough, without voice recognition or speech-to-text. It does not determine that “Person X” coughed, only that a cough occurred.
“If this can achieve the same level of accuracy without relying on a camera or additional sensors, then I think we did the job,” Ismail said. “Especially if you think about indoor environments like a hospital.”
Additionally, the system only processes enough audio to run its detection, then immediately deletes the raw sound after its analysis is complete.
What’s next
While CoughNet acts as a proof of concept, Ismail has plans for developing the system further.
In its current state, CoughNet listens to audio and deems that it is a cough. A large goal going forward is to analyze coughs for more context by examining their specific acoustic signatures and checking if the sound characterizes the cough of a specific disease (such as a flu-like illness, COVID, or bronchitis).
“We’re looking at whether some diseases have cough features or sounds that we can make use of to give some sort of probability — this person is coughing and he might have X or Y,” Ismail said.
The obvious application is to install the technology in medical waiting rooms or units to alert staff of a spike of respiratory illness in a defined area, which would in turn better inform them and allow for more efficient allocation of resources.
However, CoughNet could have direct-to-consumer applications. Ismail wants the technology to run on active noise-cancelling headphones and wearable technology to turn those devices into background helpers. For instance, the device in theory could alert someone if they are entering an area where there may be at an elevated risk of respiratory illness.
There are many avenues this technology could take, but that’s what is so exciting about it for its creators.
“There are many applications we can think of, but this is our first work in this health domain, and we’re pretty excited,” Ismail said. “These types of work have an impact on society. That’s what’s interesting about them.”
Computational simulation/modeling
Not applicable
CoughNet: A Lightweight, Low-Cost, and Energy-Efficient Multiple Coughers Detection and Identification System
31-Aug-2026