The paper introduces an extension to the D* algorithm, which is designed for real-time replanning in dynamic environments where arc costs can change during the traversal. The D* algorithm, an incremental version of A*, efficiently repairs paths as new information becomes available, but it can be inefficient when the robot's sensor data reveals significant discrepancies in the initial map. The extension focuses the repairs to minimize the total time required for initial path calculation and subsequent replanning operations. This is achieved by using a heuristic function similar to A* to propagate cost increases and focus cost reductions, ensuring that only the necessary states are expanded. The paper presents the intuition behind the algorithm, defines the necessary notation and formulation, describes the algorithm in detail, and provides empirical results comparing the performance of the Focussed D* algorithm with other algorithms. The results show that the Focussed D* algorithm outperforms the basic D* algorithm in terms of both offline and online time, making it suitable for real-time replanning in dynamic environments.The paper introduces an extension to the D* algorithm, which is designed for real-time replanning in dynamic environments where arc costs can change during the traversal. The D* algorithm, an incremental version of A*, efficiently repairs paths as new information becomes available, but it can be inefficient when the robot's sensor data reveals significant discrepancies in the initial map. The extension focuses the repairs to minimize the total time required for initial path calculation and subsequent replanning operations. This is achieved by using a heuristic function similar to A* to propagate cost increases and focus cost reductions, ensuring that only the necessary states are expanded. The paper presents the intuition behind the algorithm, defines the necessary notation and formulation, describes the algorithm in detail, and provides empirical results comparing the performance of the Focussed D* algorithm with other algorithms. The results show that the Focussed D* algorithm outperforms the basic D* algorithm in terms of both offline and online time, making it suitable for real-time replanning in dynamic environments.