At the Complex Systems Labs at IIT Delhi, research has made it possible to generate recipes using AI, tailor recipes to reduce carbon footprint leading to more sustainable food practices, create diet plans to address specific health concerns, and more. It’s all being done under a decade old area of research named Computational Gastronomy.
Computational Gastronomy is an elective that is being offered for Data Science and Computer Science students at IIT Delhi. Its champions are seeking to introduce it in other institutions as well.
Computational Gastronomy has piqued the interest of a range of experts. But what is it and what does it mean for the students of Culinary Sciences, Nutrition and Dietetics?
Computational Gastronomy is where Data Science and AI meet food. Pioneered by Ganesh Bagler, a computational researcher, the area has ventured to understand food and cooking, through the lens of structured databases. The lab creates algorithms for analysing food data, food pairings, culinary fingerprints, taste and sweetness prediction.
Work so far
The team has created various data repositories for food. RecipeDB, a structured repository of recipes from across the globe includes 1.18 lakh recipes from 74 countires while FlavorDB is a repository of flavour molecules found in food ingredients. They have also created SpiceRx, a platform for exploring the health impacts of culinary herbs and spices, DietRx, a platform for exploring the health impact of dietary ingredient, and SustainableFoodDB, a repository of carbon footprints of recipes.
The lab has created a bot named Ratatouille, named after the movie, which generates novel recipes. “You can choose ingredients and it will generate a new recipes, which will of course be informed by traditional recipes and patterns”, says Mr. Bagler.
Mr. Bagler says, with the carbon footprints data of ingredients like chilli, potato, meat and everything else available, they have built optimisation algorithm which can minimise the carbon footprint of any recipe you are generating. “That is one way to achieve sustainability. So the food we eat is tasty as well as leaves less carbon footprint”, says Mr. Bagler.
Education and careers
Computational Gastronomy has been taught as an open elective for the last five years at IIT Delhi. It is taken up by students of Computer Science, Data Science and other similar programmes. Mr. Bagler is also creating an NPTEL course for the subject which will be released next year.
Mr. Bagler says AI will disrupt Culinary Science. Since Computational gastronomy touches on anything and everything related to food recipes, sustainability, flavours, nutrition and health, it is going to be an enabler for professional chefs, nutritionists and dieticians.
Computational Gastronomy can be introduced in programmes of Culinary Science, Hospitality Management, and other related disciplines. “A full-fledged standalone course can happen down the line. The field has to become popular first. But students of these disciplines need to equip themselves with these AI tools, just the way professionals are adapting”, says Mr. Bagler.
Chef Akshay Malhotra says that Computational Gastronomy will help people become more efficient operators and better chefs. He says AI will help a great deal in decision making. Chefs read and memorise books about mother recipes and their derivative sources, which will be now easily accessible through AI.
Mr. Malhotra also says that crucial information will be on fingertips like what ingredients should be eaten together. “It’s like buying a car. We can compare recipes in terms of Sugar, Salt. Whether I should go with condensed milk or Khand. If a dish has a higher than average daily consumption of salt, they could give alternate recipes where salt is used less. It will help in pairing the right food with the right sauce, the right dip with the right chip, you can go as fine as that and make your food better”, says Mr. Malhotra.
Elaborating on what this means for nutritionists, Mr. Bagler says, rule of thumb, non personal recommendations made by nutritionists are going to be out of sync with reality. If two people have type-2 diabetes at the same level, prescribing the same standardised diet for both people doesn’t work. Different body types would react differently to it. “Wearable AI devices can help in identifying activities and patterns of the person and recommend a diet specifically for the particular individual”, sayd Mr. Bager.
Mr. Bagler says adapting to AI is not longer a choice for Chefs and Nutritionists. “AI won’t replace them. But those who refuse to use AI, will definitely be replaced by those who do”, he said.
Inception
During 2014-2015, Mr. Bagler was teaching a Complex Networks Course at IIT Jodhpur. He says one of the earliest papers in this area showed flavour networks, i.e. how ingredients shared flavour profiles. There was an existing hypothesis of how ingredients are brought together to construct recipes for western cuisines. “My immediate thought was what is true for western cuisines may not be true for Indian”, he says.
After that Mr. Bagler got academically invested in this area. For a year, Mr. Bagler used data science to understand the peculiarities of Indian cuisine and published his research which soon gained popularity. “I realised this is an opportunity to create a new area. One that no one has looked at. To see food and cooking beyond artistic endeavours, through the lens of data”, says Mr. Bagler.
Symposium on Computational Gastronomy
IIIT-Delhi is organising the sixth edition of the Symposium on Computational Gastronomy on Saturday, December, 14 2024. It will showcase recent computational gastronomy research from the institute. It will have talks by chefs, industry representatives, policy-makers, and influencers.
Chefs will give their perspective about how data driven technology can revolutions Gastronomy. Pawan Agarwal, CEO, Future Food Foundation, will talk from nutrition education perspective. Dr. Soumya Swaminathan, Chairperson, MSSRF, Chennai, will talk about using AI for improving health at ground level.
Published – December 04, 2024 08:26 pm IST
