Abstract Target tracking problems have been extensively studied for both robots and sensor networks. In this paper, we consider a target tracking problem in which a sensorless tracking robot must maintain close proximity to an unpredictably moving target. To assist the robot, a network of sensor nodes, each equipped with a binary proximity sensor, is spread through the environment. This architecture has the benefit of eliminating the need for information-rich sensors on the tracker, while supplying it with nonlocal observations of the target. However, it also introduces new complications due to the mobility of the tracker and the energy limitations of the sensor nodes. To address these issues, we present algorithms that allow both the tracker and the sensor nodes to maintain partial information about the target's location. The contribution of this work is an algorithm that manages the propagation of information across this network, making message delivery decisions on-the-fly, based on each message's informative value for the tracker. We present an implementation along with simulation results. The results show that our system achieves both good tracking precision and low energy consumption in both start-up and steady state phases of the problem, and that its performance is superior to that of earlier methods for this problem.