In sensor network applications, data gathering mechanisms, which are based on multi-hop forwarding, can be expensive in terms of energy. This limitation challenges the use of sensor networks for applications that demand a predefined operational-lifetime. To avoid this problem, using of Mobile Element (ME) as a mechanical data carrier has emerged as a promising approach. However, practical considerations such as the ME speed and route planning, sensor buffer size and data frequency generation constraints impose limits on this approach. To address these issues, we propose a natural hybrid approach that combines two approaches of ME and multi-hop forwarding. We consider the problem of determining the path of the ME, in which the length of this path is bounded by pre-determined length. This path will visit a subset of the nodes. These selected nodes will work as caching points and will aggregate the other nodes’ data. The caching point nodes are selected with the aim of reducing the energy expenditures due to multi-hop forwarding. To address this problem, we present a heuristic-based solution and compare its performance against the optimal solution. We obtain the optimal solution by providing an Integer Linear Program for this problem.