Table of contents
It is no longer a question of if Artificial Intelligence [AI] will see greater use in the food processing industry. Instead, it is simply a question of how significantly and how swiftly AI in food service is catching up.
Imagine a world where produce is grown with heightened nutritional and flavor profiles. This produce is sorted with minimal wastage and makes its way to manufacturing plants. Here, it is incorporated into recipes based on consumer preferences amid stringent food safety regulations. With end-to-end visibility and optimized inventory, the product is sold rapidly.
You reserve a table through your phone. Walk into the restaurant and browse the menu at leisure. You order without being conscious of your preferences. Until now you have only interacted with a bot. You are served by a waiter who knows your tastes and makes excellent recommendations. You leave feeling satisfied with the amazing experience and promise to return.
You no longer have to imagine. AI makes this a reality.
AI in food service is not taking over jobs. Instead, it is working with personnel to create a humane and competent source to table ecosystem.
Top Applications of AI in the food industry
While the food industry has not been the most amenable to technology in the past, AI will be the catalyst for bringing in this revolution.
AI in the food and beverage market was valued at USD 3.07 billion in 2020. This is expected to reach USD 29.94 billion by 2026 at a CAGR of over 45.77%.
Evolving consumer needs for fast, high quality, affordable, and easily accessible food options have precipitated these changes.
Sorting Fresh Produce
Sorting through produce is one of the most time-consuming processes faced by businesses that receive fresh produce. This is an incredibly time-consuming, labor-intensive job. AI helps in sorting fresh produce based on the end-use optimizes usage while maximizing resources.
For instance, sorting produce by size. Whether manufacturing french fries, chips, or hash browns, machines are currently programmed to deliver consistent products regardless of the individual characteristics of each potato. Unfortunately, this uniform processing disregards the natural variations in the size and shape of potatoes leading to food waste.
TOMRA, a company founded in Norway, develops sensor-based optical sorting solutions with machine learning capabilities. Potatoes most suitable for the end product [french fries, wedges, or crisps] are identified and separated by deploying various technologies, including cameras and near-infrared sensors. TOMRA claims their sorting and peeling solutions recover 5-10% of produce by reducing the amount thrown out.
Similarly, TOMRA also works in sorting through a fresh delivery of tomatoes, and picking out off-color ones decreases the probability of rejection by the retailer or consumer.
Deploying AI in sorting fresh produce leads to
- Improved efficiency
- Reduced waste
- Higher yield
- Reduced sorting time
- Enhanced customer satisfaction
Streamlining the Supply Chain
As food safety regulations grow stricter, companies are required to maintain compliance and transparency. AI proves to be of immense help in these situations. Being able to evaluate and observe food safety measures throughout the supply chain helps in maintaining accountability.
US food waste is about one-third of everything produced, equaling about 60 tons and $160 billion each year. While some of that waste happens in homes, a large amount is lost in the supply chain.
By 2035, Accenture predicts that AI will boost profitability by an average of 38% in the grocery industry by reducing wastage.
Symphony RetailAI, headquartered in Texas, helps in streamlining the supply chain in food service by deploying AI. Their technology is capable of providing actionable insights based on real-time data driving improved performance and profit.
Leveraging AI in supply chain management leads to:
- Transparent Pricing
- Maintaining Food Safety
- Simplified Inventory Management
- End-to-End Visibility
- Maintaining Compliance
- Reducing Wastage
With AI tracking market demand, there is end-to-end visibility in the supply chain. This ensures that the available products are optimally assigned and dispatched, avoiding wastage.
Maintaining Food Safety Compliance
Food safety is non-negotiable and AI in food service is playing an important role in this department as well. This begins with maintaining good personal hygiene in plant floors or kitchens.
Some of the common pain points in food plants and restaurant kitchens include:
- High consumption of human resource
- Low monitoring capabilities
- Manually checking and reviewing videos
A few years ago, KanKan AI signed a deal to provide an AI-powered solution for improving personal hygiene among food workers in China. As a result, both restaurants and manufacturing facilities can deploy this system.
KanKan works through a combination of technologies, such as
- Cameras to monitor workers
- Deploying facial-recognition software
- Leveraging object-recognition software
Maintainance of standard safety by using appropriate gear can be monitored based on safety laws. Upon detecting any violation, pertinent images from the recording are pulled up for review. The technology deployed by KanKan AI proves to be over 96% accurate.
Deploying AI in kitchens and manufacturing floors help in:
- Uniform Recognition
- Life Cycle Supervision
- Real-time monitoring
- Checking for sanitized hands
- Recognizing apparel [hats, aprons, hairnets, gloves, masks]
- Behavior Pose Detection
- Intelligent Enforcement
- Trash Bin Detection
- Pest and Rodent Detection
- Disinfection Detection
Cleaning processing equipment
Cleaning processing equipment in the food service industry is time-consuming and highly labor-intensive. It requires the following:
- Large amounts of water, consumables, and human resources.
- High amount of energy consumption
- Eating into the productivity of employees
- Excessive use of chemicals
- High cost to manufacturers
- Toxic to environment
To solve this, researchers at the University of Nottingham are developing a system that uses AI to reduce cleaning time and resources by about 20-40%.
The above uses a multi-sensor approach including visual and auditory systems for Self-Optimized Clean-In-Place [SOCIP] monitoring. SOCIP deploys ultraviolet, optical fluorescence imaging, and ultrasonic acoustic sensors to detect and measure food residue and microbial debris within various equipment. This technique subsequently optimizes the cleaning process. Further, this could reduce cleaning times by up to 50 percent, enabling less downtime and more productivity.
The research predicts that the system could save the UK food industry £100 million per year.
Developing New Products
Based on evolving consumer expectations, food manufacturing is changing with respect to ingredients, recipes, and the endless permutations and combinations available.
Gastrograph AI uses ML and predictive algorithms to model consumer flavor preferences. The data generated is analyzed and works toward simulating model consumer flavor preferences and predicting how they will respond to new tastes.
AI, in concert with ML, helps in categorizing the data generated into multiple demographic groups. These help companies develop new products that are in tune with the preferences of their target audience. For example, deploying AI in the food industry allows manufacturers to predict and anticipate what products will flourish before launch.
Coca-Cola began the installation of self-service soft drink fountains in various locations allowing individuals to customize their drinks. Based on the options available, customers could theoretically create hundreds of different beverages by combining different flavors to their primary beverages.
These drink fountains each dispense hundreds of different drinks a day, simultaneously generating significant data about consumer preferences.
The data demonstrated that consumers created a large number of cherry-flavored Sprite that it would do well as a standalone product. As a result, Cherry Sprite was the first product to be developed based on the insight gained.
Developing recommendation engines is a prime application for AI in food service industry. These will suggest new products and ingredient combinations that will flourish in the market. These will start with high volume, low-cost foods and work their way towards more complex, layered foods.
Today the scope of AI has evolved from simply automating. AI in food service industry is being used to understand what flavor combinations different demographics use.
For instance, Kellogg Company launched Bear Naked Custom, allowing people to make their customized granola from over 50 ingredients.
Powered by IBM’s Chef Watson, the system has analyzed thousands of possible ingredient choices. Then, the AI makes suggestions on potential ingredients to add to your granola and offers input on the taste.
However, Chef Watson does not just stop with this. The data generated through this on flavor combinations, preferences, and the number of re-orders for a particular variety creates a feedback loop. Campbell and other companies are deploying Chef Watson through Watson ads in choosing recipes.
Data analysis enables a deeper understanding and refinement of the flavors consumers love. In addition, the data generated provides valuable information on newer products and flavor combinations.
TasteMap uses Deep Learning to recommend wines based on experience, environmental facts, and taste. And with Spoonshot, you can make use of the higher-order insights to drive future forward food innovation decisions.
Higher Quality Food
An estimated 4.1 million data points are predicted to be generated every day at farms by 2050.
While this is a little way from realization, it is not too far off. There is intensive research underway on how AI and ML could help farmers grow better food by creating optimal growing conditions.
The optimal growing conditions are unique to different crops. Food computers equipped with multiple sensors collect data aiding the growth of small amounts of produce in controlled environments. The premise here is that improved flavor profiles and tolerance to external stressors could pave the way towards higher yield, maximal resources, and reduced wastage.
The data collected from these sensors can help determine the best growing conditions for different plants with specific targets using predictable outcomes. For instance, improved flavor profile in basil.
Trace Genomics extracts DNA from the soil, analyzes its microbial community, and provides AI-based recommendations for maximizing soil health and crop yield.
The discovery of unique and critical connections between growing conditions is possible due to AI.
Sentient Technologies, an AI-based company, is working on observing the effects of variables like UV light, salinity, heat, and water stress on basil.
Research from Cornell University demonstrates how a neural network was built and trained to identify brown leaf spot disease on cassava leaves with 98% accuracy. CAMP3 uses and controls wireless sensor networks to gather field images and spot plant diseases and pests early on.
Additionally, AI is being deployed in farms to identify plant diseases and pests, enhance soil health, improve nutritional quality and yield.
Front-End Applications of AI
Gartner predicts that by 2020, a whopping 85% of enterprise-customer relationships will be managed without human interactions.
Another application of AI in food service is the leveraging of AI-based virtual assistants such as chatbots to resolve queries, take orders and reservations. Virtual assistants significantly reduce the waiting time, allowing the system to take on basic questions and transfer only serious issues to customer representatives.
There is immense potential for chatbots in restaurants to enhance customer engagement and welcome a hyper-connected demographic. Hotels have begun integrating chatbots in their operational processes. This has led to a remarkable ROI especially by automating their concierge desk.
AI-based self-ordering machines lead to an enhanced customer experience by eliminating queues and reducing the waiting time. In addition, these kiosks can take orders and payments directly through integrated card readers, eliminating human interference.
These kiosks leave customers feeling more comfortable as they can peruse the menu at leisure and order without feeling self-conscious about their preferences. Integrating voice ordering with these kiosks is the next big thing to reduce physical touchpoints.
Some restaurants are going a step further by displaying nutritional information. For example, Edamam generates a real-time nutritional analysis of food recipes by extracting food entities from unstructured text using Natural Language Processing.
This reduces the load on employees, allowing them to focus on creating a great customer experience and getting to know each customer individually.
Streamlining Business Operations
AI makes it possible to
- Create and manage employee schedules
- Real-time inventory tracking
- Gain insight on
- Busiest times
- Weather and local events
- Track social media trends
This optimizes operations and frees up some of your employees’ time so they can focus on mission-critical tasks.
AI works with employees to create a great customer experience by analyzing past data to anticipate future expectations. Then, the staff can use this information in real-time, providing customers what they want even without asking, leading to loyalty and repeat customers.
The Road Ahead
Bringing end-to-end visibility, advanced sensors, and a nuanced understanding of consumer choices, AI in the food industry is here to stay.
AI brings a comprehensive understanding of context and intent that explains customer behavior, needs, expectations, and purchase decisions.
In concert with ML, AI will help drive brand loyalty, better nutrition, and reduced food wastage through data analytics.