Efficient Eats: Food Delivery Optimization and Restaurant Employee Scheduling

Efficient Eats: Food Delivery Optimization and Restaurant Employee Scheduling

Founded in 2013, Maestro Pizza quickly became the leading pizza brand in Saudi Arabia and has been the go-to choice since 2019. As part of the Daily Food Co. family, the company is renowned for its dedication to providing outstanding service and delicious pizzas, earning widespread acclaim.

Problem

In 2022, one of the key development goals set by the company was to find solutions for food delivery optimization and boost productivity. To analyze and model the business processes, Maestro Pizza approached Jaco-Ben Consulting.

Throughout the work on the Maestro Pizza case, the consultants highlighted the following needs:

  1. Measure the customer service levels with different system configurations.
  2. Implement a cost-effective method to test the impact of staffing changes and food delivery route optimization.
  3. Provide a range of likely outcomes for different demand patterns and restaurant employee scheduling.
  4. Test the impact of changing prioritization in the delivery and execution philosophy of the services provided by Maestro Pizza.
  5. Incorporate various dynamic events and system interdependencies into the analysis.

To get the data-supported results, the consultants used AnyLogic simulation software. Using GIS maps, experiments, and other features, they created a detailed model of the Maestro Pizza network and developed valuable insights for business growth.

Solution

The current Maestro Pizza business model is presented in the scheme below.

Maestro Pizza scheme of the restaurants workflow

Maestro Pizza business model

Within the system, the company employs four types of employees: kitchen staff, cashiers, delivery drivers, and part-time drivers. They handle tasks in various restaurants and delivery processes. The kitchen staff members are versatile, as they can take part in every stage and substitute for any other member. However, there are a few dedicated duties, like “cut” and “making,” that only kitchen staff members can work on.

Since kitchen human resources are not interchangeable, the key questions for food delivery optimization revolve around determining the optimal quantity of each resource and establishing restaurant employee scheduling.

To formulate an effective strategy, the consultants divided their route to a solution into two distinct phases.

Phase 1. Food delivery route optimization via simulation

The company started by importing data from the Maestro Pizza file into AnyLogic. The file contained information about schedules, priorities of different staff members, and maximum and minimum staffing in each branch. Then, the consultants added information about order generation logic.

Maestro Pizza branches models for food delivery optimization
Simulation model of all branches with a detailed view of each restaurant

The simulation in the first phase replicated the actual delivery process and provided statistics for further food delivery optimization.

Client could interact with the model by clicking on drivers, orders, or branches to access detailed information about wait times, estimated delivery times, and more. All stores are simulated simultaneously, offering information about staff and deliveries across all branches.

GIS map in AnyLogic for food delivery route optimization

GIS map with food delivery routes and dynamics

In AnyLogic, GIS maps are at your disposal. They help you plan routes on actual roads, considering past traffic patterns that affect drivers' speed. It's handy for food delivery route optimization. You build the model with the road network as it is in real life, export it from AnyLogic, and then run it even without an internet connection.

Phase 2. Experiments for store workflow optimization

In the second phase of the food delivery optimization strategy building, the focus was on optimizing the current workflows within the stores using the results of the first phase.

For the analysis, the company ran a complex parameter variation experiment based on the model they built in the first phase. This helped to explore different scenarios and see the impacts of minor changes on the restaurants’ workflows.

Utilizing the input data provided by Maestro Pizza, Jaco-Ben Consulting developed a model that considers variations in restaurant employee scheduling, order intervals, labor efficiency, and target service times.

Setup for experiment in AnyLogic to determine proper restaurant employee scheduling

Experiment setup for restaurant employee scheduling

After the experiment calculations are finished, there is an option to download a detailed summary of the experiment. It downloads as a .txt file, but you can open it as an Excel table. That’s why AnyLogic is convenient not only for consultants but also for end clients. All the data is accessible within the model.

Considering the uncertainty of future analyses and potential changes, there's no need to overly focus on incorporating charts directly into the model. The richness of information, encompassing all possible outcomes, is already embedded in the model's log file. This way, the model and its data were presented to the client, enabling them to explore various scenarios on the spot.

Results

Minimum number of drivers and branch staff by type

Based on the simulation results, it was highlighted that on Sundays, the client needs five fixed drivers and one floating, whereas on Saturdays, the number of fixed drivers remains the same, but they need three floating ones.

Chart with number of drivers by type, per shift, per day of week

Maestro Pizza divided branches into groups so that floating drivers could effectively switch between deliveries.

Correlation of kitchen staff required with cashiers and drivers

Chart with comparisong of kitchen staff vs cashiers

Having more cashiers has little impact on food delivery optimization; the outcome stays flat. However, if the restaurant adds a delivery driver, it will take less time to prepare the orders. If there are not enough drivers, kitchen members must deliver the food themselves. Unfortunately, this results in queues in stores.

Chart with comparisong of kitchen staff vs delivery drivers

So, for food delivery optimization, it is better to have an additional driver than an additional cashier.

For more details, please check out the PDF case presentation by Jaco-Ben Consulting or watch the video from the AnyLogic Conference 2023.


同类案例研究

更多案例研究

获取包含行业案例研究的手册(英文)

下载