“It smells like apple pie!”, the robot says. This statement w | Benefit Daily

“It smells like apple pie!”, the robot says.

This statement won't sound ridiculous in the future, as the machines are already learning how to smell. Google researchers are training neural networks with a new technique to predict how a molecule smells based on its chemical structure.

It's not that easy to make a robot tell one odor from another. One atom or bond, “and you can go from roses to rotten eggs,” says Wiltschko, who led the Google research team for the project. Many other scientists tried to figure out how to make it possible, but Wiltschko’s team took a different approach. They used something called a graph neural network, or GNN. Most machine-learning algorithms require information to come in a rectangular grid. But not all information fits into that format. GNNs can look at graphs, like networks of friends on social media sites or networks of academic citations from journals. They could be used to predict who your next friends on social media might be. In this case, the GNN could process the structure of each molecule and understand that in one molecule, a carbon atom was five atoms away from a nitrogen atom, for example.

The Google team used a set of nearly 5,000 molecules from perfumers who have expert noses and carefully matched each molecule with descriptions like “woody,” “jasmine,” or “sweet.” The researchers used about two-thirds of the data set to train the network, then tested whether it could predict the scents of the remaining molecules. It worked.

However, this research doesn’t tell us much about mixtures or combinations of scents, which can radically alter how we perceive single molecules. But figuring out what properties or patterns lead molecules to smell a certain way would be a huge advance for the field.

Alexei Koulakov, a researcher at Cold Spring Harbor Laboratory, says that the project is valuable for introducing thousands of new molecules into the smell data sets, which are often relatively small, and that this data “could form the basis for improvements of this and other algorithms in the future.”

There is still a long way to go.