Fig . 7 : A screenshot of the post-facto dashboard providing an executive summary analysis of physical distancing violations . The sliders at the top allow the user to define a violation by selecting distance and interaction time parameters . Based on the selected parameters the violations are summarized using the violation percentage , severity of violation and average violation time . The pie-chart gives a breakdown of the violations for each geofence specifying a different region . Finally , the scatter plot shows an interaction plot for each person which is segmented into four regions based on the threshold parameters set for distance and interaction time .
The dashboard in Figure 8 provides details about the violations in different locations and its distribution over different times of the day . It is important to note , that each of the plot can be filtered using day of the week and location to get in-depth detail insight about the situation . The location-based analysis chart not only provides the user insight about the violation percentages but also its comparison total person visiting the location .
Bar / Line chart on the left indicate the customer arrival (%) and violation (%) by day of the week . There is a positive correlation between number of customer and violation (%). For example , Tuesday , Wednesday & Thursday are high-volume days at SAS café , and we see higher physical distancing violation on these 3 days . On the right , we have a tree map showing the areas of violation . The size of tree map indicates the location where people most of their time in SAS café while colors indicate violation (%). Red color locations have higher violation (%) while the green color locations have lower violation (%). For example , salad bar is the area where people spend most of the time , but dessert is the area with most per-capita violation (%). This makes sense as people are continuously moving in salad area while people tend to gather around desert and usually take longer to decide what they would like to choose . We can also see violation by time of the day at the bottom . There are few counter-intuitive trends here . We would have expected
- 14 - November 2020