The objective of the fifth week During this week （ the last week ） , our group has integrated the project. In the indoor positioning part, the software we use the clustering algorithm. The algorithm first randomly assigns each observation point to one of the K categories and then calculates the mean of each category. Next, the algorithm reassigns each point to the category whose mean is closest to the point data before recalculating the mean. Repeat this step until you don't need to redistribute it. There are some variations of this algorithm. There are several ways to initialize the cluster for "seeding". Here, we randomly assign each data to a group and then calculate the mean of those groups. This causes the average between the initially generated groups to be close, resulting in more repeating steps later. Since the armor was not perfect, we decided to replace the armor with the clothes and pasted the vibration module onto the clothes and put the Arduino board