Great writeup. This focuses heavily on the fragrance/perception side of digital olfaction, but there's a whole other dimension here: detection. Think of it as speaker vs microphone — most of the hype (Osmo, Givaudan) is on generating and designing scents, but detecting and identifying chemical signatures in real-world environments is an equally hard problem.
The article nails the core challenge: the structure-odor relationship is messy and there's no "RGB of smell." We're hitting the same wall from the detection side. Getting to parts-per-trillion sensitivity is one thing; making sense of complex, noisy molecular signatures in the wild is another. The DREAM challenge benchmarks are encouraging, but the gap between lab conditions and real-world deployment is enormous.
Curious if anyone else is working on the detection and safety rather than fragrance. Feels underrepresented relative to the urgency.
Great writeup. This focuses heavily on the fragrance/perception side of digital olfaction, but there's a whole other dimension here: detection. Think of it as speaker vs microphone — most of the hype (Osmo, Givaudan) is on generating and designing scents, but detecting and identifying chemical signatures in real-world environments is an equally hard problem.
The article nails the core challenge: the structure-odor relationship is messy and there's no "RGB of smell." We're hitting the same wall from the detection side. Getting to parts-per-trillion sensitivity is one thing; making sense of complex, noisy molecular signatures in the wild is another. The DREAM challenge benchmarks are encouraging, but the gap between lab conditions and real-world deployment is enormous.
Curious if anyone else is working on the detection and safety rather than fragrance. Feels underrepresented relative to the urgency.
Disclosure: I work in the scent detection space.