Can we meet the rising global demand for food? For centuries, farmers and food industry stakeholders tried to answer this question using technology. Some argued that automation, machinery, and mass production are the keys to sustaining a growing population. Others have argued that dealing with the complexity of food-related data requires advanced data science and computer science techniques such as artificial intelligence and machine learning. In this blog post, I will explore the latter.

The population of the earth is increasing. Humans keep up with the growing demands of food availability by exploiting plan and animal resources through agriculture, forestry, and fishing. Growing crops, raising and breeding animals, and harvesting timber and other plants, animals or animal products are necessary for the human race.

The food business is growing. Agriculture and foodstuff products are produced to feed the world’s growing population. Canola, wheat, soybeans, meat, livestock, and animal feed are a few examples of foodstuff products. Food and beverage products are exported to feed people of this world. Domestic and overseas export requires strong business relationships with farmers, processors, and agronomic experts across multiple countries.

Artificial Intelligence, also known as AI, is a field of computer science that deals with intelligent machines. Machine learning and deep learning are two of the most commonly used algorithms in the field of AI. These models learn from data and are used by people, companies and government agencies to make predictions. The goal of machine learning is to explore the infinite probability space in order to come up with the most suitable solution to any problem. Today, machine learning models are being developed to deal with the complexity and variety of data in the food industry.

The question arises: “Can AI and machine learning help us deal with the growing food demand?” To answer this question, let’s look at AI ideas, AI applications, and AI research.

AI Ideas for Food Industry

Artificial intelligence projects often require brainstorming sessions. There are many AI and machine learning opportunities to consider and coming up with new ideas can be challenging. At Produvia, we share our visions on how to apply AI in the food industry.

Achieving Zero Hunger

It’s time to re-think how we grow, share and consume food. Agriculture, forestry, and fisheries have the ability to provide nutritious food for the world. Today, our soils, freshwater, oceans, forests, and biodiversity are rapidly declining and degrading. Climate change is putting even more pressure on natural resources that we depend on, increasing disasters such as droughts, hurricanes, and floods. Poor food security is causing millions of children to be stunted in development due to severe malnutrition. A major change is needed if we are to nourish 800+ million people who are hungry today. An additional 2+ billion people are expected to be undernourished by 2050.

It’s time to address world hunger using AI and machine learning. It’s possible to analyze growing, manufacturing, distribution, and consumption data to make intelligent predictions and recommendations for food industry stakeholders. We can build an AI-platform that not only understands supply and demand but can also adapt to changing population needs and desires. By striving towards zero hunger, we meet one of the United Nations’ Sustainable Development Goals in order to achieve a better and more sustainable future for all (UN, 2015).

Responding to Global Demand for Foodstuffs

Today, the global population exceeds seven billion people. This figure is forecast to reach more than nine billion by 2050. As economic development brings wealth and prosperity to more areas of the world, the growing population is widening the gap between food supply and demand. Responding intelligently to foodstuff demand is required to meet global challenges. The goal is to deliver safe foodstuff products from around the world in the most efficient matter.

Forecasting global demands, delivering safe foodstuff products can be done using machine learning and deep learning. It’s possible to use time-series data to build forecasting models. It’s also possible to analyze unstructured data to build prediction and recommender models in order to ensure that global demands are met. These models can be integrated into existing business processes to ensure that food and beverage stakeholders make informed decisions.

Ensuring Stable Food Supply

Securing grains, oilseeds (wheat, corn, soybeans) and other raw ingredients through a series of networks and partnerships are often necessary to ensure the availability of stable foods, livestock, and aquaculture feed, cooking oils and many other foodstuffs worldwide. From production through to distribution, maintaining logistic efficiency and safety in the food supply is necessary to ensure that the food supply remains stable through time.

Creating a stable supply of foodstuff sources can be accomplished using AI and machine learning. Generative models and genetic programming can be used to explore food market conditions never considered before. Recommender and prediction models can be developed to analyze hundreds of thousands of market factors. These models can drive decision making to improve food supply stability.

AI Applications in Food Industry

There are many companies and organizations that already incorporate machine learning, deep learning, and AI into food and beverage products and services. Here are our favorite AI applications from the food and beverage industries.

  • Dodo Pizza generates pizza recipes using deep learning technologies by combining two recurrent neural networks (RNNs). Dodo Pizza AI learned to find non-obvious connections between pizza ingredients, understand how to pair ingredients and how the presence of each influences the combinations of others. Dodo Pizza also open-sourced their code online (Github, 2019)
  • Gastrograph AI uses machine learning and AI to understand consumer’s sensory perception of flavor and predict consumer preference of food and beverages (Gastrograph AI About, 2019)
  • Whisk uses deep learning and natural language processing to map the world’s food ingredients, properties (nutrition, perishability, flavor, categories) and food purchase options in order to provide hyper-relevant advertising and customized personalizations (Whisk FAQ, 2019)
  • Tastry uses AI, machine learning, analytical chemistry, flavor preferences to provide consumer product recommendations. The company provides retailers with science-based suggestions for product development, inventory purchase and direct-to-consumer recommendation (Tastry Press, 2018)
  • Vinify uses machine learning to advise customers on wines adapted to their tastes (Vinify Asset, 2019) while TasteMap uses deep learning to recommend wine according to experience, taste, and environmental factors (TasteMap Retail, 2019)
  • Edamam uses natural language processing for the extraction of food entities from unstructured text in order to provide the nutritional analysis (Edamam Developer Documentation, 2019)
  • Pingwell is exploring the use of computer vision and machine learning algorithms to deliver contextual information to consumers and retailers in the grocery/pharmacy space (Pingwell, 2019)
  • Sure is applying natural language processing and machine learning to better understand user’s needs and navigate them to the right businesses in the overcrowded food scene (Sure Blog, 2016)
  • Instacart uses machine learning to predict real-time availability of grocery items (Instacart Blog, 2018); Instacart leverages machine learning, representation learning, and image similarity search to sort items for users to “buy again” and recommend items for users while they shop (Instacart Blog, 2017); Instacart uses deep learning and emojis to sort shopping lists (Instacart Blog, 2017); Instacart developed machine learning models to predict the distribution of time expected for any given shopper (Instacart Blog, 2017); Instacart uses machine learning to predict the number of orders for each delivery window for a given store (Instacart Blog, 2018); Instacart built regression and machine learning models to manage variance and ensure on-time deliveries (Instacart Blog, 2018); Instacart leverages machine learning models to predict the likelihood of claim and cancelation of any particular hours in order to improve supply planning (Instacart Blog, 2018).

AI Research in Food Industry

At Produvia, we read the latest machine learning research, so you don’t have to. Here are the latest AI research projects related to the food and beverage industry.

AI Ideas for Food Industry

We brainstormed three artificial intelligence ideas which can be applied for packaging and food service industries:

  1. Material Identification —identify polymers, plastics, and microplastics automatically using computer vision and machine learning to save hundreds of hours in sorting and recycling
  2. Product Development Recommendation System — identify polymers or plastics most suitable for new products using machine learning to save hundreds of hours in research and development
  3. Packaging Recommendation System — identify packaging options best suited for new products using machine learning and recommendation systems to save hundreds of hours in product development

Conclusion

Today, there are many use cases for AI and machine learning in the food industry. Some of the world’s leading startups and enterprises are already using machine learning and deep learning in their operations.

Are you engaged in the production, accumulation, marketing, manufacturing, and processing of food resources and products? If so, let’s chat! Visit us at produvia.com to start a conversation around the topic of AI, food or beverages.

Next Step

Interested to solve food problems?

Schedule a call with Slava Kurilyak, Founder/CEO at Produvia.