Esma Birisci, a Mizzou Engineering doctoral alumna, and Ron McGarvey, assistant professor of Industrial and Manufacturing Systems Engineering, have been working on the problem of planning food production schedules for all-you-care-to-eat facilities for years now. And they’ve recently added a new twist to their formula, allowing for reusable leftovers to factor into the equation.
Birisci and McGarvey recently published “Optimal production planning utilizing leftovers for an all-you-care-to-eat food service operation,” in the Journal of Cleaner Production. This paper follows the path set by their previous research published in 2016, which considered the production of items that cannot be retained as leftovers, such as fried foods, and factored in the environmental cost , stated in terms of carbon dioxide emissions, embodied in the food item’s creation.
“What we did here was extend it out to the full range of foods they make at Campus Dining services,” McGarvey said. “That includes some fried food, which can’t be retained as leftovers, along with many other items that can be stored and saved for future use.”
Using a combination of optimization models and data analytics techniques, the researchers identified the relationship between production planning decisions and food waste and unsatisfied customer demands.
“This full range of food items makes the production-planning model more complicated, because every day you make these decisions, you need to look back at the retained excess from the previous day. For the fried stuff, everything that isn’t sold is thrown away. For most other food items, though, you don’t necessarily lose overproduction if you can carry it forward as leftovers,” McGarvey explained.
Birisci, who now is a faculty member at the University of Uludag in her native Turkey, and McGarvey again utilized data from Mizzou’s Campus Dining Services (CDS) to come up with models that allows organizations to factor in the what variables are most important to their production in order to optimize results.
These variables include whether a food service operation is more worried about total food waste weight or the environmental factors that go into producing the wasted items. For example, beef has a higher environmental cost than most vegetables, given all that goes into feeding and raising a cow, and the emissions the cow itself produces.
What the research also found was altering production for certain items during certain meals had a wider impact than less targeted production adjustments. Were CDS more concerned with overall weight of the waste, adjusting the meat during lunch on Wednesday, the vegetable during Wednesday dinner and Sunday dinner, and the fried pork, poultry or fish option during Monday lunch optimized production.
On the other hand, if CDS was more concerned with environmental impacts, adjusting the Wednesday lunch meat selection, Friday’s dinner vegetable choice and Sunday’s dinner meat had the biggest impact.
“She had to identify the optimal production levels for different items on different days of the week and different meals,” McGarvey said. “She tuned that to two different objectives. You can think about your objective being minimizing the weight of the food you throw out. … And an alternative would be to think of the greenhouse gas emissions embodied in the food.
“Within that, she identified which are the key meals and key food types you’d want to focus on to achieve the minimum amount of waste under each objective.”
For example, it’s better to overproduce reusable food resources such as vegetables on a Monday rather than a Friday, given that demand is likely lower over the weekend in a campus dining hall than during the week.
Birisci and McGarvey hope to collaborate with CDS and its inventory management software vendor to incorporate this logic into this commercial software, potentially allowing any food service operator to perform similar analyses and identify which meal and food type combinations would generate the greatest impact on differing waste objectives.