IBM is one of the biggest tech companies in the world, and its vast resources have allowed it to come up with some pretty outlandish inventions. Back in April, we reported on the company's controversial human resource AI, which possesses the ability to analyze employee habits and work performance to determine when they are about to quit with 95 percent accuracy.

Now, the company has come up with something slightly less Orwellian but still somewhat unusual: an electric tongue that can be used to "taste test" various liquids for hazardous waste or chemicals. This device, appropriately called the "Hypertaste," can perform in-depth liquid analysis across a "variety of compounds" with minimal equipment necessary; making it ideal for scientists who lack access to high-end laboratories.

IBM calls the Hypertaste an "AI-assisted e-tongue for fast and portable fingerprinting of complex liquids," and that's an apt summary of the tech's functionality. Hypertaste relies on an array of "cross-sensitive" electrochemical sensors made up of electrode pairs. These pairs can "taste" molecule combinations present in a given liquid and assign it a specific voltage signal.

When these sensors have performed this process for all liquid molecules, a specific "fingerprint" for the liquid can be determined. According to IBM, the entire process of fingerprinting and chemical identification takes "less than a minute," which is relatively impressive for a portable device such as this. The results of a given analysis can be quickly viewed via a dedicated mobile app.

In terms of the possible use cases for Hypertaste, IBM offers quite a few examples in its announcement blog post. With Hypertaste, government agencies could perform on-the-fly water quality testing in remote locations. Alternatively, beverage or food companies could easily verify the product being shipped by suppliers indeed contains the ingredients it's supposed to; without the need to send anything off to a lab to wait for lengthy testing and results.

Hypertaste is, of course, just a prototype for now and it will be some time before it ever rolls out for enterprise use. However, the idea is interesting, and we're looking forward to seeing how IBM improves it in the future.