Abstract:
For many years, navigation systems for outdoor use have been available. It's only recently become popular for indoor use. Thanks to technologies such as Bluetooth and Wi-Fi, public spaces such as shopping malls have developed solutions that make it easier for visitors to navigate the area. At the same time, a method known as crowd-sourced localization has become popular.
Indoor occupancy detection systems have grown in popularity over the last decade, owing to devices such as Bluetooth Low-Energy beacons, which are inexpensive and effective.
The goal of this thesis is to use machine learning techniques to help users navigate to their workplace, office equipment, conference rooms, and other areas in an office environment. This could make it easier for staff to work more efficiently while also saving time. The gathering of RSSI data and (x, y) location coordinates from fixed beacons is researched and implemented using fingerprinting techniques for indoor positioning. A literature review of various machine learning models that are appropriate to our system is conducted. The chosen models were tested and compared using performance indicators such as the CDF curve, MAE, RSME, and Euclidean distance error to see how well they performed. For maintenance and people tracking in a working area, this system can be tracked in real-time.
With the increasing complexity of large-scale building interior layouts and the increasing amount of time people spend indoors, providing users with effective and comfortable indoor navigation path planning is becoming increasingly important. We conclude that Dijkstra is one of the best techniques to determine the shortest path between two points based on our research, analysis, and comparison of other path-planning algorithms.
