As the ‘Industry 4.0’ has been proposed, the Internet of Things (IoT) has been widely used in more and more fields, and the higher demand for data acquisition efficiency. Unfortunately, the format, type, and access methods of data sources in different areas are diverse, making low efficiency of data acquisition. To address this issue, based on the data acquisition middleware of IoT (DAQ-Middleware), a novel High-speed Data Acquisition Algorithm of IoT (HDAA) is proposed. The algorithm includes parallel data acquisition algorithm and acquisition efficiency optimization heuristic method. According to the characteristics of different data sources and data interfaces, the mapping relationship is established between data sources and data interfaces, which improves the data acquisition efficiency of IoT. In the practical application and simulation environment, the data acquisition efficiency of HDAA is analyzed and compared with other algorithms. The results show that HDAA greatly improves the efficiency of data acquisition. In particular, the advantage of the algorithm is more obvious in the case of more data sources and less data interfaces.