Abstract: In sparse wireless sensor network, neighboring nodes are far away from each other. The requirement of direct communication among sensor nodes in traditional static data collection methods causes large time-delay and high energy consumption. In order to solve above problem, the paper proposed a data collection method of wireless sensor networks based on mobile robots. The method firstly takes a mobile robot as cluster head node which is nearest to a static sensor node, then judges node moving according to comparison result between average remainder energy of neighboring nodes detected by mobile robot in a period and average remainder energy of all nodes in the network. The mobile robot moves according to random moving strategy in controlled range. When the mobile robot arrives at a new position, the sensor node renews route and selects a new mobile robot as cluster head. The simulation results showed that the method can reduce time-delay of data transmission and energy consumption of sensor nodes greatly compared with traditional static data collection methods.