By Rajat Bhageria, Founder/CEO of Chef Robotics
Key Takeaways:
- Traditional automation has been challenging for high-mix food manufacturing due to frequent product changeovers and the need for flexibility. Low-mix producers can easily use fixed automation, but manufacturers with multiple SKUs and changing recipes struggle to implement cost-effective automated solutions.
- Human labor remains valuable in food production due to its adaptability to changing conditions, ingredient variations, and recipe modifications. However, hiring and retaining skilled workers is increasingly difficult and expensive, creating a need for more flexible automation solutions.
- AI-enabled robotics represent an emerging solution that bridges the gap between traditional automation and human labor. Using a “see-think-act” model with advanced sensors and machine learning, these robotic systems can adapt to varying conditions, handle multiple SKUs, and potentially provide a more efficient and consistent approach to food manufacturing.
While the worst of the Covid-era labor shortages are behind the industry, labor has been and continues to be a major challenge for food manufacturers. Always grappling with rising costs and impacts to the bottom line, we’ve seen growing interest and adoption of automation across the primary food packaging space. While high volume, low SKU mix producers have historically invested in traditional automation to meet demand (think of your major brands like Lay’s, Kraft Heinz, etc.), their high-mix counterparts dealing with dozens or hundreds of SKUs have had limited options to effectively automate beyond the basics of conveyors and sealers.
Why have high-mix producers had limited options to automate?
The answer lies with flexibility and changeovers. For production lines making the same SKU day after day, low mix manufacturers can get custom automation per SKU and just run it day in and day out. Additionally, equipment tear down and setup can be effectively managed around shift schedules. Traditional automation, like dispensers and depositors can work efficiently in this type of environment.
The reality, however, for many food manufacturers across fresh foods, direct-to-consumer, frozen foods, and contract manufacturing is much different. Many production lines experience several changeovers within a shift, often having to handle different tray types, or even changing between meals as varied as salads, burritos, and sandwiches within the same shift. The guardrails to maintain process stability to make traditional automation work reliably often cannot be set in this type of environment. Traditional automation has not been flexible enough to be successful.
How should my team evaluate automation solutions?
While there have been limits to automation in high-mix food manufacturing, leaders looking to drive improvements in their operations have several options in today’s market:
- Evaluate the feasibility of dispensers/depositors
- Continue to leverage labor
- Get the best of both with AI-enabled robotics
When to leverage dispensers and depositors on the line
If your operation is low mix, these solutions are cost-effective and reliable options to drive process improvements and savings within your facility. They tend to be high-performance once set up and commissioned, and are “easy to hire” with readily available options on the market.
For some higher mix manufacturing operations, they may still be a cost-effective solution if one or two longer runs of meals can be allocated to a particular line, reducing changeover overhead to a minimum.
With that said, flexibility tends to be more limited, and for many high-mix operations, these solutions will not work well. As a thought experiment, if you have 300 ingredients, you’d likely need hundreds of custom-made depositors. Each one would have low utilization and spend most of its time idle, making a case for positive ROI difficult to justify. Further, if your marketing or culinary team changes the recipe or formula, you may need additional expensive changes made to the hardware. These factors should be considered when evaluating the lifetime cost of the equipment.
The benefits of keeping humans in the loop
The benefit of human operators is clear — they provide large flexibility to a range of conditions. Variables like changes in ingredient characteristics, different trays, or even conveyor speeds are relatively easy for humans to adapt to and handle. If a recipe change occurs, a new SKU can be run with simple training before getting started. This ability to adapt is why labor is so prevalent in fresh and frozen prepared meals, meal kit producers, and airline and patient trays.
This adaptiveness, however, comes with a clear cost. Hiring and retaining talent is hard in today’s market, and hiring costs or lost production output due to understaffing quickly add up. Maintaining quality with churn is also a constant struggle for operations managers.
AI-enabled robotics: the emerging option for flexible production
To break out of the labor paradigm and stay competitive, food manufacturers are beginning to look towards new forms of robotics enabled by modern advances in AI. These solutions bring the best of both worlds — they’re highly flexible, reliable, and can maintain strong performance across multiple shifts. In some cases, onboarding a new SKU can even be done by a simple web app.
How is this possible? Let’s talk about the see-think-act model of robotics:
- See: With a variety of sensors ranging from cameras to highly precise depth sensors, robotics systems can detect items in their environment and build a picture of the world around it.
- Think: With state-of-the-art AI and Machine Learning, robotics solutions can interpret these environmental signals and determine how to best move and operate to accomplish a task, even in novel conditions.
- Act: After planning what to do, these AI-enabled robotic solutions are able to flexibly execute and accomplish the needed task.
Prior automation was predominantly only performing the “act” step and doing the exact same action over and over again, with no or limited sensors to understand the environment. With more sensors, computer vision models allow robotic systems to see items like trays on conveyors or ingredients in a bin and behave differently for every task.
These capabilities make this new generation of automation robust to variance — different or varying ingredients, different trays, new portion sizes, changes in conveyor speeds, or even entirely new conveyors are no longer blockers to high performance.
These recent advances in robotics and AI are going to enable a step change in efficiency, quality, and operational excellence for food manufacturers. This shift gives early adopters the opportunity to win over customers by being able to deliver superior products at competitive prices.
So what’s the verdict for established high-mix manufacturers looking to improve their operations? More than likely, it’s a mix of all three. If feasible, try to leverage depositors for longer runs with stable production throughout the year. Then use AI-enabled robots to tackle a range of ingredient deposits previously incapable of being automated. And finally for anything that neither can handle well, use humans as the underlying technologies continue to improve.
