With the wide and rapid development of wireless sensor networks (WSNs), current WSNs have become more and more complex with lager capacity, which need emerging technologies to support. For example, artificial intelligence (AI) technology is required to deal with huge data transmission and parameter optimization. Big data and data mining technique is required to collect a large amount of data and explore those data to find hidden significance. In this paper, we essentially investigate the way of application of abovementioned technologies including AI, big data, and data mining in heterogeneous sensor networks. We focus on the following three application scenarios including resource scheduling, unmanned air vehicle (UAV)-assisted data collection, and three-dimensional spatial path autonomous deployment of heterogeneous sensor networks. We discuss the various aspects involved in each scenario. For different scenarios, AI algorithms such as proximal policy optimization (PPO), pointer network (PN) and multi-agent deep deterministic policy gradient (MADDPG) are used to optimize the sensor networks. Through a case study, we reduce the energy consumption of WSNs and verify the superiority of AI technology and data mining in heterogeneous sensor networks. I believe that this study will inspire future research on heterogeneous sensor networks.