The Raspberry Pi, since its launch, has not only been a real commercial success, it has also enabled the creation of a large number of projects. Some are very curious, some are very helpful, and a few others are really curious. Well, one user created two watches with two Raspberry pie defraud artificial intelligencewho have the capacity to listen and feel.
If you search the Internet, there are many interesting projects. There are different irrigation systems for gardens and small orchards, video surveillance systems, doorbells with cameras and even a smart mirror.
We have also seen different projects related to artificial intelligence based on the Raspberry Pi. What we didn’t expect is that someone would develop a watch capable of listening and, above all, detecting odors. The latter can be very interesting, especially if combined with a type of air freshener when bad odors are detected.
They create watches that can hear and feel
It’s not new that there are watches based on Raspberry Pi. We have different projects that can tell us the time, but we have seen some healthy rivalry within the community to develop more and more clocks complex.
Designer and developer Max Björverud has created two rather complex and, to some extent, non-functional watches. It has integrated artificial intelligence into these watches to estimate the time based on smell and sound.
Björverud developed these two watches which operate completely independently. The one that detects odors incorporates a multi-channel gas sensor. On the other hand, the one that detects sounds integrates a microphone. Both watches should improve over time.
They don’t give time as usual, they estimate it based on the stimuli they receive and what they have learned so far. Both watches feature matrix systems displaying hours, minutes and seconds.
They include a bottom raster panel that generates shapes based on metrics from the latest collected data. Regarding precision, the creator insisted that uncertainty is the goal of this project. Björverud emphasizes that his project wants to help raise awareness about artificial intelligence.
An interesting aspect is that each of the watches uses an independent Raspberry Pi and they are different models. The Listening Clock uses a Raspberry Pi 3 B+, while the Smelling Clock uses a Raspberry Pi 4 B. Both projects, as time passes and they collect data, can indicate time more accurately.
For the artificial intelligence part, he used an open source Python library called Scikit-learn. This library has many tools suitable for machine learning projects. Raster plots are powered by the open source C++ tool openFrameworks.
We are talking about perhaps one of the most curious projects we have seen to date. It does not have light sensors, which leaves the “nose” and “ear” entirely free to interpret what time it is. That it interprets the time of day based on these parameters is rare and one needs to know how accurate it can be.