The amazingness of the human nose and its olfactory ability is undeniable.
It is estimated that each nose contains approximately 400 scent receptors, and these have the capability to distinguish up to one trillion distinct aromas.
It is a daunting challenge to recreate the amount of sensory knowledge found in scientific tools.
Thanks to advances in artificial intelligence (AI), the recent electronic noses - sensors that can detect and report particular smells - have been improving rapidly in terms of speed and accuracy.
Those who support them claim that they can revolutionize food safety.
Salmonella and E. Coli possess a type of personality that is unique to them, according to Prof Raz Jelinek, the co-creator of Sensifi, an e-nose, and a professor of chemistry at Ben Gurion University of the Negev, Israel. He adds that each of these potentially life-threatening foodborne bacteria has its own distinct electrical signal.
The Israeli company, e-noses, has created electrodes featuring carbon nanoparticles which are capable of detecting volatile organic compounds (VOC) released by bacteria through smell.
Varying strains of bacteria trigger unique VOC fingerprints in the Sensifi device, triggering varying electrical signals. AI-driven software logs the signal and compares it to its accumulating collection of data, with its user notification capability.
Since its launch this year, Sensifi has been striving to change the face of the fight against food-borne infections. Modi Peled, its CEO, says that producers of foodstuff typically have to send samples off for lab testing, with results requiring several days to be returned.
In comparison to other options, Sensifi's e-noses can be utilized at the food firms themselves, with results in under sixty minutes. The company has not yet released the cost of their machines, describing them as being "economically friendly". Instead, the bulk of their profits are expected to come from subscription costs.
Mr Peled states that testing methods in the food industry have stayed similar for the last four to five decades, until recently when AI began to be employed in the assessment of this sector.
Look into other narratives about AI
It is estimated that 48 million Americans, or one in six people, suffer from a foodborne illness annually in the US. These cases result in 128,000 hospitalisations and prove fatal for 3,000 individuals. Food poisoning is a major global issue.
It is estimated that in the UK there are 2.4 million cases of food poisoning annually, with an associated 180 fatalities.
Mr Peled commented that while it is commonly believed that meat, poultry, and fish are the primary culprits, romaine lettuce has in fact been the primary contributor to food-related deaths in the United States over the past five to ten years.
As the food industry becomes more industrialized, it is increasingly vulnerable to contamination from pathogens.
At NTT Data Business Solutions, a German company, they devised a unique method for training the AI behind their developing e-nose - coffee.
For an experiment, technicians took three days to place instant coffee granules near the AI's receptors. The AI had to distinguish one of three options - a quality coffee, a poor coffee (with vinegar added), and no coffee at all.
Adrian Kostrz, the innovation manager of the company, states that an aroma is not merely a gas, but rather a one-of-a-kind blend of gases. Additionally, he adds that often, there are minute distinctions in the manner an item smells.
NTT has equipped a 3D-printed plastic model of a human nose with its sensors, and is providing its AI with coffee and other food items in order to teach it the smell of freshness and quality, which will be the AI's "reference value of an odour".
It is proposed that NTT's e-nose can not only detect any illness-causing pathogens in the environment, but also determine the freshness of food. This could help supermarkets and restaurants know which food to sell first if it does not come with an expiry date.
Mr Kostrz states that an awareness of the reference value of a smell could be advantageous to the food industry in terms of optimizing production techniques, storing, collection and other operations.
Although the output of the most recent e-nose models is proficient, AI specialists predict that there will not be a noteworthy need for them due to the expense, which may deter food companies.
Vincent Peters, the founder and chief designer of US-based AI research firm Inheritance AI, comments that when speaking of deploying a global system of little sensors, from choosing to storage to delivery, one must contemplate how this will affect the business model.
Do you think the business will be able to effectively deploy and maintain the technology? Will the supply chain management take on the responsibility of implementing it? Is there a potential for profit?
Kjell Carlisson, an AI specialist from San Francisco-based Domino Data Lab, mentions that the e-noses require thorough adjustment to each site in which they're employed. He goes on to point out that this is a very challenging undertaking in a sector notoriously unreceptive of new innovations.
Despite the scepticism, some entrepreneurs remain undeterred.
Scentian Bio of New Zealand claims to have copied the antennae of insects to create its "biosensors", utilizing insect proteins for their scent sensors.
Andrew Kralicek, the originator and chief technology officer of the company, asserts that the biotechnology they have created allows for their sensors to be "thousands of times more proficient than a canine's olfactory apparatus".
He claims that this biosensor-based tech can be put to many uses, for instance, food and flavour quality control, food pathogen detection, non-invasive quick disease diagnosis, sustainable farming, and environmental and wellness surveillance.
Comments