• Everyone hates lines. In recent times, however, this everpresent problem has been exacerbated by the COVID pandemic, were standing in a queue with too many people can not only be a pain but a public health hazard.
  • Recent covid stats say that in the US there have been increases in new cases up to 41,844 in a day alone. espite this, we've seen a movement to reopen more and more restaurants due to economic concerns. Something that we want to encourage while maintaining safety. Harris County had 274 overcrowding complaints this weekend due to reopening, most likely due to the nature of how visits to restaurants are distributed. Due to rush hour, people are mostly visiting these restaurants at times where other people are simultaneously visiting, thereby increasing the population density at certain times.
  • While places like Rice do a great job at maintaining social distance, public facilties with long lines at these rush hour time periods compromise this and pose a health risk to eager patrons. To ensure that patrons can still go to their favorite public places while minimizing the public health risk and attempting to redistribute the population densities, we developed our solution: Qtime.

Our Solution

  • QTime is a web widget that allows users to check-in and out using QR codes of restaurants and public spaces that expect lines or queues. Our dynamic Bayesian prediction model congregates the data observed from users into a list of safety levels depending on the time of day. Our widget outputs the current safety level based on the device's current time.
  • The incentive system is set up to reduce the points earned from visiting during "red" or packed times, and increase points earned from visiting during ‘green’ or safe times, so customers are incentivized to visit when conditions are safer.
  • With our unique machine-learning solution, our idea is to be able to dynamically update the model with respect to the number of people who check-in throughout the day, so that the model can improve in its prediction accuracy with more training data.
Share this project: