Construction operations include many complex tasks and dependencies among resources. Traditionally, collecting operational data involves several manual steps to convert time-lapse photos or videos into productivity data. These manual steps are time-consuming, labor-intensive, and prone to errors. Several technology advances have made new approaches possible for data collection and analysis. Many researchers have experimented with Global Positioning System (GPS) and Radio Frequency Identification (RFID) technology to track the resources on site. By integrating pre-specified boundaries which are associated with the site layout or activities, engineers are able to track the current status of the resources or the current activity in which these resources are engaged. However, these positioning data can be further manipulated to provide more detailed information regarding construction operations such as resource bottlenecks, utilization, and balance. This paper presents an on-going research that analyzes the dynamic geometrical data of resources to extract construction operational information. Preliminary results are presented to demonstrate the feasibility of automatically converting real-time positional data of tracked resources once the geometrical patterns (operational model and strategies) have been identified. A conversion model representing the geometrical patterns is used to transform the input, the positioning data ObjectID(x, y, z, time) collected from RFID or GPS at regular time intervals, into information for productivity analyses. A case study of concrete drainage pipe installation is used to verify and validate the concept. Further discussion will be presented on precisions, lessons learned, and limitations of the relevant technologies.