Monitoring the productivity of trenchless construction processes, such as Pilot Tube Microtunneling (PTMT) installations, is necessary for proactive control of construction operations. For example, engineers can correlate construction operation parameters (e.g., forces) with contextual information (e.g., soil type) for identifying factors influencing the PTMT productivity. Awareness of these correlations can help engineers to select and control the equipment accordingly. Such correlations, however, are hidden in large amounts of field data. Tedious manual data collection and processing cannot capture and analyze details of PTMT workflows. This paper presents an automated data collection and interpretation approach for supporting detailed PTMT productivity analysis. This approach uses a data logger to record the hydraulic pressures of equipment used during the PTMT automatically. A two-step pattern recognition method can detect time-series patterns of the hydraulic pressure and identify cycles of operations in three stages of the PTMT: 1) pilot tube installation; 2) casing installation; and 3) product pipe installation. The first step uses an Artificial Neural Network (ANN) to classify the time-series as belonging to a certain stage of PTMT. The second step uses an Adaptive Anomaly Detection Algorithm (AADA) to split the time-series into sections corresponding to operational cycles. A case study demonstrates that this automated approach can reliably recognize operational cycles of construction equipment in PTMT workflows.