5 minutes read.
Food and Beverage (F&B) is a continuously growing market driven by population and middle-class raise, dynamically changing consumer requirements, food safety and quality concerns, sustainability efforts and digitalisation opportunities. Companies are permanently facing growing competition and are challenged by the changing global trends.
Typical competitive strategy of F&B companies is a combination between differentiation and cost leadership. They constantly run cost synergies, savings programs and supply chain (SCM) optimisation. Through lean approach in complex manufacturing with short runs, F&B companies drive operational excellence, efficiency improvements; implement waste reduction, run sustainability efforts, and strive for optimised logistics. Furthermore, premium quality assurance comes via rigorous raw materials and ingredients selection, processes control for perfect taste and innovation of flavours and packaging. Additional improvement opportunities are coming through the application of newest technologies.
The secret ingredient - Artificial Intelligence (AI)
AI has been gaining significant momentum over the last few years, with many of the players from F&B industry actively investing in various applications across the value chain. This emerging technology is helping F&B companies optimise production and maintenance, apply predictive analytics, increase transparency, engage in digital marketing etc. Here are a few possible applications in the food industry where AI can provide an enormous benefit:
Ensuring food quality and limiting food waste
Quality control starts from raw material intake to semi-finished goods to packaging lines and logistics. Typically, detailed raw material quality characteristics are not available along the entire value chain. Very often the reasons for required rework or scrap of the final product are known only after the manual quality inspection. Real time quality checks, using AI, could result in significant quality improvements and reduction of the cost of bad quality. Sampling and quality checks during production could be automated and performed in real time using various sensors and spectral analysis techniques, driven by AI. In addition, vision systems based on AI can replace human visual inspection and contribute to productivity gains and increased reliability.
Many of the raw materials, as well as end products in F&B are perishable. Therefore, companies need to manage sensitive processes and cold chain, including cooling rooms and refrigeration units. Machine learning and AI technologies like reinforcement learning provide a great opportunity for optimisation of these processes across multiple dimensions. Such technologies can help simultaneously reduce energy consumption while keeping the perishable materials safe.
Machine learning and AI can also help retail chains automate inventory management, through automatic notifications to reorganise or restock the shelves dynamically, based on the real time customers demand.
Finally, using AI for automatic inspection of packages and logistic materials before dispatch provides a great opportunity for reduction of customer complaints and claims.
Growing better food
In the traditional, as well as in the controlled environment agriculture (CEA) facilities, AI could help farmers grow better food by creating optimal growing conditions. Leverage of AI to monitor impact of light intensity and distance from growing herbs, soil condition, salinity, heat, and water demand could support efficient operations, optimization of labor activities and development of optimal growing strategies (“recipes” for plant growth).
AI vision system and sensing technologies can also be used to detect plant diseases and growth abnormalities (too high, too wide plant structure, leaves colour etc.), to improve soil health by triggering automatic fertilisation, to control environment based on the current conditions (temperature, humidity) and predictions, like weather forecast, for the amount of available natural light and water.
One of the most time-consuming processes in any facility that receives fresh food products is sorting. For example, sorting apples by similar size and quality level, or potatoes by size, to decide which ones should be made into French fries. In these scenarios, AI vision systems could dramatically reduce complexity of the sorting equipment and simplify the sorting process.
Ensuring people safety and highest hygiene levels
In a food plant, retail or in a catering kitchen, people safety and highest level of personal hygiene are necessary to ensure compliance with food safety requirements and regulations. Both food safety recalls and health and safety incidents including people are very costly and can cause brand reputation damage. AI-powered solutions for improving safety and personal hygiene typically use cameras to monitor critical processes and operations. Such systems can automatically detect liquid spills, for example, and warn people of hazards caused by them. Vision systems can also constantly monitor if employees are wearing safety equipment, required by food safety regulations, and raise alarms in the case of violations.
Assisting in predictive maintenance
Today, typical maintenance plan is based on predefined time intervals and not on the current asset condition or the way it is being used. This results in unplanned downtimes, inefficiency and considerable labor cost. By instantly processing conditions of all the assets, AI systems could help predict upcoming operational issues and support planning of maintenance actions to prevent downtime. Combination of maintenance expert knowledge and data about the actual machine operation conditions enable development of AI systems that could learn process specific maintenance patterns to minimise downtime, extend machine life, improve throughput and Overall Equipment Effectiveness (OEE).
Optimising the cleaning process
Food production requires very high levels of hygiene, to ensure food quality and safety. Therefore, F&B companies are relentlessly applying stringent cleaning procedures to ensure hygiene of equipment and the environment where food is produced. However, in many cases, cleaning procedures are activated although there is no need for cleaning at a particular time. The problem is in the lack of information on whether the cleaning is required or not, and therefore it is safer to always activate the cleaning process.
By detecting (through machine vision or other sensory techniques) state of equipment cleanness, AI systems can learn cleaning patterns suited to particular production runs, thus reducing the cleaning time and saving precious water and energy.
Detecting the food related fraud
Many consumers lost confidence in what they eat every day following reoccurring scandals (melamine scandal in China with baby milk, horse meet scandal in the UK or mislabeled seafood and fraudulent honey in US). The question “Is the food we eat safe?” is being asked more and more today, as food fraud becomes a growing problem. According to the World Customs Organization (WCO), the counterfeit food industry is estimated to generate over $49 billion a year.
AI systems combined with transparent, block-chain based, digital tracking technologies can help fight the counterfeit food activities. AI can check composition of raw materials and ingredients to ensure authenticity, while block-chain technologies provide necessary infrastructure for traceability through the entire food supply chain.
Described use cases and applications are only few from many opportunities where AI can contribute to sustainable food production. The Createsi team is very passionate and involved in these topics. We would love to hear from you. Please contact us via the web site, e-mail, or through social media if you have any comments, ideas, or if you wish to talk about a concrete problem and possible solutions using machine learning and AI technologies.