What does your nose tell you? We researched, we tried, we failed. Perhaps not so much failed, as didn’t find a suitable use case.
Maybe you can do better?
If you can include scent into an environment, individualized, for short or for long term and also based on some triggers, what would it be you’d find interesting and what would you like to experience and test?
Tell us what you think and help us to come up with a really super interesting study here. Just put your idea in a comment below!
A while ago, we heard about the possibility to test a device which enables us to produce customized scents. We found this super interesting, as we had several use cases in mind already where smell and scent were the missing pieces in the overall story. Our initial thought was: Wow, amazing! Let’s do it and bring some more senses into play!
Including this device into our Smart Dressing Room, so that we could surround the customer with an individualized scent either due to his user profile and preferences or matching the scent with the clothes he brought into the cabin, like a fresh sea breeze, sun lotion or coconut for swim suits and bathing gear or like fresh cut grass or wood scent for outdoor clothes.
As this set up would require a good ventilation inside of the cabin, we dismissed it again, as an excess of multiple different scents might be more counterproductive than stimulating after a certain time….
Another idea we had was more about training an AI on what smell and scent means for the individual. We loved the approach saying:
We are training an AI to understand human senses better by learning associations of a certain scent in the first and second level. At the end the AI will build up knowledge on what scent associations are for the individual – and also in general – in combination with colors, experiences, sound, light or pictures. This training set up was supposed to be like a game, so that the users are kept in the process to answer more questions…. this setup should be using voice / NLP enabling the user to answer and let the AI talk to the user as well – making it a bit more “emotional”.
This would be similar to what google did with using pictures as captcha verifications to let humans select “all pictures with cars” to train an AI in the backend…. The goal is using this information for individual customer engagement. Smell can be first step and next we go further by training the AI on human associations with sound, touch, taste, ….
Nice approach, but no: Smell and scent are areas where training of an AI is almost impossible, as it’s so hard to provide the right data to verify scents against, and it’s also a matter of individual associations .
Can you do better?
Tell us what you think and help us maybe to come up with a really super interesting study here. Simply drop a comment below!