The technology behind electronic noses (also known as “e-noses”) and digital olfaction is generally defined as the digital capture of aromas and fragrances – a process that can be applied to many different use cases and industries. Similar to a human’s sense of smell, digital olfaction essentially mimics the process by which our brains identify and differentiate between odors.
But to fully understand how this works, we must first define and break down the different odor sensing technologies. Let’s dive into the two main instrument-based techniques used to analyze odors – traditional e-noses and bio-based solutions.
Defining traditional e-noses
Traditional e-noses based on metal oxide sensors (MOS) have been used longer than bio-based sensors for odor sensing – although not widely and with limited success, partly due to their lower number of sensitive sensors (typically less than 16). Their sensitivity lies at the parts per million (ppm) levels, limiting its applications, and most traditional e-noses come with relatively simple and basic software solutions, requiring the use of more advanced, external software to gain deeper insights. However, their main challenge is that they are known to be prone to drift and are highly sensitive to humidity, preventing their effective use in food and beverage applications because of the interference of the water signal. In these situations, labs may try to ‘dry’ a signal by removing the water bias, but this can add too much time to the sample analysis for operators, making timely and accurate analysis difficult.
Other traditional e-noses are based on conductive polymers (CPs). However, their production process is not as mature as that of MOS, and they exhibit slow desorption kinetics, making each measurement relatively lengthy and cumbersome.
Defining bio-based sensors
Bio-based technologies are a relatively new innovation within the world of odor detection devices. They are based on biomimicry and result from an in-depth understanding of the olfaction mechanism. In these devices, dozens of sensors act as biological molecules similar to the olfactory receptors (ORs) covering a person’s nasal epithelium within the nasal cavity. These can range from short peptides derived from ORs, to real full-size ORs, to olfactory neurons expressing recombinant ORs on their surface.
Aryballe’s approach
Our digital olfaction technology is a combination of biochemistry and advanced optics, grafted on silicon sensors and intending to mimic the human sense of smell. Here’s an overview of how our digital olfaction process works:
- Odor analysis takes place at the silicon sensor level. The silicon chip is grafted with an array of patented biosensors, which have unique interactions with odor molecules.
- These interactions are visualized as dots that represent each biosensor. The lighter in color the dot is, the more intense the interaction between volatile organic compounds (VOCs) and biosensors. The difference in color intensity across the array of dots defines the pattern
- This pattern is in fact the odor fingerprint of each smell, and it is visualized as a radar chart. Each peak on the radar chart represents one type of biosensor. We call this representation a “signature.”
- Captured odors are then stored into a database, where they can be compared with previously recorded odors. Essentially, this is pattern recognition, sometimes helped by machine learning algorithms. In addition to recording the odor fingerprint, we also capture the odor intensity, which can provide important insights for further learnings.
To extend the life of our sensors, we’ve been able to decrease the minimum amount of time during which VOCs are in contact with them, reducing the risk of cumulative pollution. To achieve this, we work with an optical technology (MZI) that allows us to visualize reaction kinetics in real time, and we’ve designed peptides that interact with VOCs in a transient manner. We’ve also made the decision to conduct testing in the gas phase, which enables us to quickly remove VOCs from our sensors.
Aryballe’s sensors can be used at any point in the supply chain or manufacturing process to ensure batch-to-batch consistency from as early on as raw material sourcing to assessing the final product.
