Real-time prediction and management of food product demand

Abstract

A real-time buffer manager system that calculates optimal food buffer levels, for both completed products and product components, based on real-time counts of restaurant patrons throughout a restaurant's property and the estimated time for them to arrive at a food ordering station. The real-time buffer manager employs a computer vision system, running a series of 2D image processing techniques that detect and track vehicles and people in several camera views. Patron counts are fed from the computer vision system into a queuing model that estimates when each patron will arrive at an ordering station. Thus, instead of analyzing historical sales data, the buffer manager according to the present invention electronically performs direct measurement of probable future demand, and electronically predicts, in real-time, what the future food product demand will be in a predetermined time (e.g., 3-5 minutes) immediately following the direct measurement of the demand.

Description

BACKGROUND [0001] 1. Field of the Invention [0002] The present invention generally relates to the restaurant industry, and, more particularly, to a system and method of real-time electronic prediction and management of food product demand, especially in quick-service restaurant industry. [0003] 2. Description of Related Art [0004] The quick-service (or fast food) restaurant industry's primary value proposition is speed-of-service—i.e., how quickly the restaurant can deliver a complete meal after a customer has placed an order. Quick-service restaurant operations are built upon the concept of preparing a limited menu of food product before customers place their orders. By preparing food ahead of time and keeping it warm in a holding buffer, restaurant employees can quickly grab food product from the buffer, bag it, and hand it to a customer. This faster speed-of-service enables quick-service restaurants to serve many more customers during busy mealtimes than a traditional sit-down re...

AI Technical Summary

Benefits of technology

This system enables restaurants to prepare the right amount of food at the right time, reducing waste, improving speed-of-service, and enhancing profitability by accurately managing food production and quality.

Problems solved by technology

Unfortunately, it's difficult for managers to know how many consumer orders they will receive over the next few minutes, so they are forced to buffer “extra” product, just to be safe.

The buffer management problem is fundamentally a trade-off between the quality of the restaurant's food and the speed with which the restaurant can serve customers.

Despite its importance to the industry, restaurant managers can do little more than make “educated guesses” as they lack one critical piece of information—when will the consumers arrive?

However, the current approach does not allow restaurant resources to be managed on a real-time basis because the variability of the correlation between past and future demand events is too large.

In other words, historic information does not allow restaurant managers to know with confidence the demand that their restaurant will see over the next several minutes; however, restaurant performance (speed and quality) would benefit significantly if product demand could be predicted accurately within the secondary shelf life of the restaurant's food products.

The current approach suffers because it infers future demand from historic demand, rather than taking a direct measurement of future demand.

The current approach becomes progressively less accurate as the time window shrinks.

Further, current “inventory management systems” are inadequate to the needs of the restaurant industry because they do not account for the-limited secondary shelf life of food products.

Unfortunately, large variances leave the restaurant vulnerable to over-production and, in practice, provide little more than rough production guidelines.

The area of management of food processing and food production facilities, such as the preparation of frozen dinners, differs significantly from the production problems in quick-service restaurants because it is not a real-time management of food processing and food production.

This is especially problematic for a serial queue, like a drive-thru where every customer in line must add the food production time to his or her wait time.

Under-production can seriously damage the restaurant's profitability by reducing the number of customers served during peak meal times.

(2) Avoid over-production because over-production reduces the restaurant's food quality and increases wastage, as food product spends too much time in the bin.

If the food product's bin time exceeds the secondary shelf life, then it must be wasted.

This method is open to a significant number of incorrect guesses, which not only waste the food product (e.g., unused sandwiches or cheeseburgers), but also consume valuable production time that was allocated to making a product that no one used.

As discussed before, a fundamental limitation of the current approach is that the analysis of historical data only infers a range of probable future demand.

Images