How Ajinomatrix is digitizing the measurements of the senses of taste and smell for the food industry?
Founder Institute HQ Palo Alto journalist Dustin Betz interviewed Ajinomatrix.org Founder and CEO François Wayenberg. Here is the transcript.
Dustin Betz: Introduce yourself – speak a little about your own background, professional and/or personal experiences and motivations that led you to founding Ajinomatrix.
François Wayenberg: I’m graduated from the university of Brussels, economist, and industrial collaborator at the AI lab at the polytechnics institute there (ULB). My initial career involved working for Mitsui, a Japanese trading company est. 1620AD, as an European food dept manager with extremely organized and stringent approach to food quality for export in Japan, and in large part educating the suppliers in Europe to be able to export their produce in Japan. This later allowed me to even collaborate with NASA through what I had learned at Mitsui.
When I worked at Mitsui, I explored what was feasible as most advanced with Deloitte and we made a plan to digitize the senses of taste and smell for the food industry.
This really kept me going for a long time and got me thinking especially:
Developing the project at Microsoft speech recognition L&H incubator Flanders Language Valley
Meeting Joel Bellenson who just had made the cover of Wired magazine, in London
Dustin Betz: Next, by way of introduction, I’ll ask you to give us your quick elevator pitch.
François Wayenberg: AJX digitizes the measurements of the senses of taste and smell for the food industry through an AI open source B2B app, allowing a food company of any size to implement AI and sensory digitization within their facilities, either for QC or product development.
It solves the problem that nowadays even in the 21st century, the digitization of the senses for smell and taste is still a very artisanal process even in medium to large companies – and even in majors, the digitization level of facilities is not complete or generalized .
Dustin Betz: For those sitting far enough removed from the food sciences industry, they may be totally unaware of the level of testing and data analysis, from the test kitchen to the lab bench, all the things that go into any any big food industry player’s decision-making around the products that they ultimately put on the supermarket shelves.
Some may have heard of a concept of ‘taste testers’ or have some inkling, but I get the sense that this is a pretty impenetrably industry for most consumers—so please explain some of the scope of the complexity here, in the ‘problems’ that you are solving for?
François Wayenberg: The players in the industry have to make decisions on how to organize the tests in order to measure the taste and aroma – they can do it internally or delegate it to an external lab. Obviously, inviting competent tasters as mouths and noses is not only very troublesome during the covid pandemic, but it already is normally: it’s costly, not so easy to organize and it takes considerable amounts of time. How does it work? People as tasters give notes on different attributes, on a scale from 1 to 10 for instance, and the results are being encoded. This is, in addition to some lab equipment and maybe a few sensors, how the data is currently being recorded.
But how this is being recored is generally happening in a still very old fashion way – in such a way that the different experiments are not standardized from one to the next, and as a consequence, the results are not being comparable.
Dustin Betz: Why Ajinomatrix / why digitization of sensory data as a key solutions for addressing these challenges?
François Wayenberg: Well, the problem then is that because of the organization of how these tests are being recorded, is that there is no modern data analysis possible on the results of the tests, and what the players in the industry generally have in their hands are a large set of dark data.
Ajinomatrix brings in an easy framework in order to have these data standardized and to help implement data mining and AI in order to make sense of the accumulated data, in order to not only interpret it in a classical data mining way, but to overcome limitations of classical sensory science in terms of data interpretability.
Dustin Betz: I understand your software already interfaces with these emerging sensors such as e-noses or e-mouths—what little I have heard about those new kinds of sensory technologies sounds very interesting, and promisingly aligned with your own tech—are there other key features that we haven’t yet talked about / that you want to highlight here?
François Wayenberg: It’s a delicate matter or issue: we have for instance a giant in flavor and fragrance coming to us asking openly how they could use our technology to basically get rid of all of their panels. Their ideal would be to simply work without those.
Read the complete transcript and watch the discussion between Dustin Betz and François Wayenberg at https://agtechfoodtech.com
AgReads® is a website dedicated to the AgTech and FoodTech space. We provide news, opinions and analysis on AgriFood Tech companies from around the world.