Goal of the Project

The aim was to create a privacy-preserving restaurant queue monitoring system based on which an accurate queue length and waiting time could be displayed for users. Additionally the idea was to predict trends affecting the number of customers, such as the lunch hour or university lecture schedules.

The premise for the system was that it would help customers choose an ideal time to visit the restaurant. The system enables customers to plan around congested times and helps to decide whether or not there is time to eat, for example between lectures.

Results

Our system successfully achieved it's main goal of showing queue lengths in real-time. It correctly detects and counts people standing in line. The use of a 2D LiDAR sensor ensures user privacy is maintained across the system.

We built a clean and intuitive user interface which clearly displays information about the restaurant queue. The website is available at https://galilei-queue.tt.utu.fi/. See the video below for a view of the page.

Our work provides a basis for future implementations of queue counting in various different settings.

Future Developments

While the current system already works effectively, some future improvements could enhance it's performance.

Firstly, the current system relies on a single 2D LiDAR sensor. However, a better field of vision and consequently an increased accuracy of people detection would be achieved with multiple sensors. Secondly, the system would benefit from enhancements focusing on optimization of real-time operation as well as from robust and scalable algorithms for modelling and prediction.

Videos