Design and Development of Web-Mobile Application for Housing Project Management Using KNN for Prediction

Project management in housing development is essential to ensure timely completion, budget efficiency, and market alignment. However, many small to medium sized property developers still use manual systems, causing inefficiencies in monitoring, documentation, and sales planning. PT Bakti Luhur Abadi is one such company that still relies on Microsoft Excel for recording project progress and housing unit sales. This study aims to develop an integrated project management system equipped with a sales prediction feature using the K-Nearest Neighbors (KNN) algorithm. The goal is to improve operational efficiency, streamline decision making, and support strategic sales forecasting. The system was developed using the Waterfall method, comprising requirement analysis, system design, implementation, and testing. A key novelty of this research is the dual platform implementation web for administrators and mobile for directors and field teams enabling real time access, structured documentation, and effective communication. The KNN algorithm was tested with 30 test data and 114 training data using K values of 3, 5, and 7. The best result was achieved at K = 7 with an accuracy of 86.7%. Functional validation using black-box testing confirmed all web and mobile features operated as expected. In conclusion, the proposed application effectively automates project management and enables accurate sales prediction. It provides practical benefits for small and medium-scale property developers by increasing efficiency, improving internal coordination, and supporting data driven planning through an accessible and intelligent solution.