Scientists are trying to decode farts with an AI algorithm  

Preparing technology that will analyze toilet sounds

Yes, you read that right. Maia Gatlin, an aerospace engineer at the Georgia Tech Research Institute (GTRI), and mechanical engineering student David Ancalle are currently working to uncover the secrets behind the acoustics of urination, bowel gas, and diarrhea.

More specifically, their scientific team is training an Artificial Intelligence (AI) to be able to recognize and analyze the sound of any phenomenon that someone might encounter in the toilet. No paper has yet been published with their research, but the initial findings suggest that monitoring our…gases could significantly benefit public health.

So wanting to study all possible sounds, the two engineers built a device filled with pumps and tubes capable of replicating the physics – and sounds – of the human body and gut. This device is called ‘Synthetic Human Acoustic Reproduction Testing’ and you can see it in the image below.

With a machine at their disposal that can realistically reproduce every toilet sound, the team turned their attention elsewhere. In particular, he is now focusing on an artificial intelligence algorithm that he hopes will one day be able to detect deadly diseases such as cholera and potentially stop an epidemic before it gets too big. The goal is to combine the machine learning model with inexpensive sensors that can be used in areas prone to diarrheal disease outbreaks.

To begin building the AI, the scientists viewed countless publicly available videos and recordings of human gas and diarrhea, analyzing the frequency spectrum of each. Then, they used the SHART engine for even more simulations.

The reason scientists think toilet sounds can help:

Audio is easier to analyze remotely than a video or the information a patient can convey to their doctor. At the same time, it is a more discreet way than a normal medical examination. So the team realized that an inexpensive device and an artificial intelligence algorithm could extract from the toilet all the information necessary to detect a related disease.

As Gatlin points out, right now diarrheal diseases like cholera are rampant in countries like Haiti, killing 500,000 children a year. In fact, they are the third leading cause of child mortality worldwide.

Therefore, enhancing the detection of these diseases will certainly aid in treatment, and may potentially prevent future outbreaks as well.

You can learn more about the whole project by clicking here.