Improving Operational Efficiencies with Virtual Copies of a Restaurant

Our client, a leading restaurant chain, wanted a better way to estimate order volume and fulfillment times at their nationwide locations. While their restaurants were already using dashboards to monitor real-time data from drive-through lanes, this setup didn’t provide a clear view of future traffic or order times. That’s when they began exploring the use of digital twin models.

As the name suggests, digital twins are virtual representations that mirror real-world systems. In this case, the goal was to create 3D models of their drive-throughs. These models would enable more intuitive monitoring and pave the way for predictive analytics, such as forecasting future service times.

Our digital twin model was designed to resemble a real restaurant as closely as possible. It captured real-time updates from the client’s time-series database via WebSocket connections and dynamically synchronized the 3D model with the real-world situation, a key requirement for a live model.

We made sure to detail the model with accurate interior and exterior features of the restaurant, and we enhanced the visual polish of the model with smooth simulation of vehicle movement through the drive-through lane(s), anchored to real-time traffic data. The model itself leveraged three.js for 3D animations and FastAPI for backend services.

Because our client also wanted the digital twin model to convey predictive analytics, we combined key drive-through and point-of-sale data to provide proof-of-concept predictive models capable of estimating order volume and fulfillment times.

We created a functional and engaging demonstration of digital twin capabilities for the client and provided documentation on how the client’s existing data infrastructure could feed digital twin models. We also studied multiple industry options for creating digital twins, such as AWS IoT TwinMaker, Azure Digital Twins, and Bentley iTwin.

Furthermore, we collected and shared insights into the operational processes of a restaurant location, which we collected through various interviews with franchisees and their teams. Altogether, the project helped lay the foundation for more informed decision-making and highlighted future opportunities for the client.