The continuing advancements in artificial intelligence (AI) are set to usher in a new wave of capabilities for food preparation and manufacturing. These advances will help address some of the thorniest problems faced by the industry and pave the way for a step change in efficiency.
Why does this matter?
To understand why these new technologies and capabilities are so critical, here are a few statistics:
- Food spoilage. Per the UN Food Waste Index report, over a trillion dollars a year of food is wasted, and 8-10% of global emissions are tied to this waste. Producing this wasted food requires nearly 30% of global agricultural land.
- Labor shortages. While the worst of the COVID-era labor shortages are behind us, the United States is facing a labor gap. By 2033, an expected 1.9 million manufacturing jobs will be unfilled in the United States. Historically, American manufacturers overcame labor shortages and rising labor costs with offshoring. Given the perishability of food, however, many food manufacturers have avoided this option but are now evaluating offshoring due to ongoing hiring challenges.
- Rising costs. Inflation has eaten into the purchasing power of the consumer. The United States experienced an 11.2% increase in food prices in 2022 and a 5% rise in 2023.
If we want to keep costs low for the consumer while reducing the environmental burden of the industry, the industry needs new approaches to operating. AI and other technologies provide that path.
How to leverage AI for food manufacturing operations
The best starting point for implementing AI will depend on your organizational needs, but for food manufacturers, there are several areas that leaders should explore.
Uncovering business insights from the production line to find inefficiencies. Modern enterprise resource planning (ERP) systems are making more data about daily operations available. To track production, manufacturers leverage an increasing number of connected devices, from networked scales to Internet of Things (IoT) sensors to track production conditions such as temperature.
However, much of that data remains siloed, unanalyzed, and in unstructured forms. With AI, teams can quickly and effectively leverage their existing data to get insights into production. Operations leaders can identify the drivers of unplanned downtime or the conditions required to maximize overall throughput and boost efficiency.
Importantly, leaders can also see real-time snapshots of their facilities to understand if things are going to plan and make adjustments throughout the day. By understanding in depth what is happening, inefficiencies can be reduced to help keep overall costs down.
Gain unprecedented quality control capabilities even faster than before. Inspection systems today are far more capable than those of even a few years ago. Camera systems with computer vision capabilities help flag issues such as bad pieces of lettuce or contaminants such as plastic film in food as a meal is being made. On fast-moving production lines, inspection systems capture missed meal components before these potential factory escapes make it out the door to customers.
By collecting images of every meal and pairing them with data from meal check-weighers or other scales, quality teams can gain a detailed understanding of where process deviations might be occurring and quickly implement improvements or retraining based on these insights. If a customer complaint or health issue does occur, teams have a strong record to reference to understand how the problem might have occurred or if other units were also affected at the same time.
The big constraint in the past was that these models needed thousands of labeled data points which is of course cumbersome to collect. Importantly, new AI models can be run with very little training data instead of the thousands of samples previously required, allowing a much faster time to first value for these systems.
By leveraging these technologies, manufacturers can boost customer satisfaction and avoid or limit costly food recalls.
Accurately forecast raw material and ingredient quantities. Especially for fresh food producers (given perishability timelines), properly ordering and preparing ingredients is critical to meeting demand without over-purchasing. Unfortunately, doing this well is quite difficult, particularly due to the seasonality of the industry.
Forward-thinking manufacturers can use AI to analyze their past order volumes to make better predictions of how much they will need to produce and adjust ordering and staffing as appropriate to meet that demand. Especially for perishable ingredients, far more accurate order volumes can be made to reduce the amount of wastage that occurs.
Over-portioning, or giveaway, is also common in the industry, and AI can help identify sources of yield loss and the proper amount of food to be prepared for a given batch size. Food manufacturers can begin to cut down on unnecessary food waste.
AI-powered meal assembly as a form of labor augmentation with the next wave of robotics. For many food manufacturers, especially those handling many SKUs, traditional automation has not been a compelling option due to the lack of flexibility. With advances in AI-enabled robotics, that flexibility challenge is beginning to be solved. By leveraging cameras and other sensors, these systems develop an understanding of their environment, and AI enables them to handle the variation expected on food manufacturing lines.
Variations in ingredients, day-to-day changes in their conditions, different conveyors or line speeds, and diverse containers are no longer barriers to automation. In fact, configuring a specific ingredient or portion size can in many cases be done through a simple app.
As a result, many teams are now evaluating how they can leverage these new technologies and augment their human workforce. They see robotics as a way to mitigate hiring and retention challenges, while driving down costs. As an added benefit, these systems remove chronic over-portioning and drive yield improvements.
In addition to the reliability and precision these systems can bring, robotic systems also generate large amounts of data useful for generating other operational insights.
Palletizing and material handling robotics for an efficient warehouse. Warehouse robotics continue to propagate the market as one of the early success stories of the new wave of AI enabled robotics. These autonomous mobile robots (AMRs) create sophisticated mappings of your warehouse and navigate between many different points, allowing for a wide range of paths in moving pallets and other goods across facilities. Many warehouse operators successfully augment their labor force to boost output by leveraging these platforms.
Excitingly, AI is also enabling material handling capabilities with more flexible, mixed SKU palletizing. With the ability to see, think, and vary action, palletizing robotic solutions today easily swap to palletizing a new SKU coming down the line. These robots also sort multiple SKUs together on the same pallet, planning out how to best stack the next set of items. Now high mix facilities with many SKUs are adopting palletizing robotic solutions previously limited to a low mix, high volume production.
Between these new palletizing solutions and AMRs, AI-enabled robotics are powering the modern warehouse and logistics operation.
In summary
For food manufacturers, advances in AI are creating both new ways to understand the production line, and new ways of meal assembly and fulfillment. By embracing these new technologies, teams can combat waste and inefficiencies in their facilities while future-proofing their operations to labor constraints. Ultimately, teams who are quick to adopt these powerful new tools will reduce costs and build a strategic advantage over their peers.
.png)
