Moving objects continuously generate large amounts of spatio-temporal data. Maintaining and querying the location information of moving objects are a crucial problem. This paper focuses on the query problems of indexing structure for moving objects. The main focus of query processing is retrieval of approximate future movement information about moving objects that may be unavailable. First, a VPMR-tree is proposed for indexing the moving objects in road networks. Second, exponential smoothing method for predicting the result of queries that refer to the future is implemented. The experimental results have shown that VPMR-tree achieves a better performance for query evaluation than TPR-tree.