10:15 - 12:15
 An electronic olfaction system rivalling the speed of smell in mice.
Thu-S10-005
Presented by: Michael Schmuker
Nik Dennler 1, 2, Damien Drix 1, Tom P. A. Warner 3, Shavika Rastogi 1, 2, André van Schaik 2, Andreas T. Schäfer 3, 4Michael Schmuker 1
1 School of Physics, Engineering, and Computer Science, University of Hertfordshire, Hatfield, UK, 2 International Centre for Neuromorphic Systems, Western Sydney University, Sydney, Australia, 3 Sensory Circuits and Neurotechnology Laboratory, Francis Crick Institute, London, London, UK, 4 Department of Neuroscience, Physiology and Pharmacology, University College London, London, UK
Olfaction helps animals navigate their environment to find food, escape predators, locate mating partners. In natural environments, odorants are usually dispersed by turbulent processes, and encounters with odour cues are usually brief and occur in a highly intermittent fashion. Odour-guided navigation therefore requires animals to be capable of detecting and recognising odours during these brief encounters with low latency and high temporal precision [1]. Indeed, it has been shown that Drosophila is capable of decoding odours within milliseconds [2], and mice can access and behaviourally report temporal features of odour plumes at timescales up to 40Hz, much faster than their sniffing rate [3].
Technical solutions to odour-guided navigation have high potential in e.g., disaster management, environmental monitoring, and security, but still lag dramatically behind the capabilities of animals. Inspired by fast olfaction in animals, we have developed a miniature accelerated electronic olfaction system based on metal-oxide sensors. We evaluated the system in a high-temporal-precision setup that was recently used in mice [3] and found that it can recognise odour pulses within milliseconds. It can match, and even exceed, the capability of mice to decode temporal features of rapidly fluctuating stimuli. The highly integrated system can be mounted on small robots for olfactory navigation experiments, e.g., in a "robotic twin" setup, where the effectivity of biological and non-biological navigation strategies are assessed in turbulent but controlled environments, while accessing information about source location encoded in turbulent-induced fluctuations of odour concentration [4].

References

[1] David, Kennedy, Ludlow. Nature 303, 804–806 (1983).
[2] Szyszka, Gerkin, Galizia, Smith. Proc National Acad Sci 111, 16925–16930 (2014).
[3] Ackels, et al.. Nature 593, 558–563 (2021).
[4] Schmuker, Bahr, Huerta. Sensors Actuators B Chem 235, 636–646 (2016).