14 POSTURE SEQUENCE ANALYSIS

Many work activities cannot be adequately represented by a single posture. In these cases, the body's evolution over time must be described by recording several successive postures that represent a task, a work cycle, or a significant part of an activity.

Ergoniza Intelligent supports posture sequences in Multiple Postures mode. In this mode, the human model does not represent only one body configuration at a single point in time, but an editable animation made up of postures placed at different points on a timeline. Instantaneous and time-dependent analyses can be applied to this sequence, including cumulative exertion load, changes in postural discomfort, and the assessment of postures sustained during an activity.

Ergoniza Intelligent can create posture sequences using several procedures. The most appropriate method depends on the type of task, the available information, and the required degree of accuracy.

The most direct way to create an animation is to build it manually from the graphical model. The user can place the model in an initial posture, record it, move forward in time, change the posture, and record it again, repeating the process as many times as necessary.

Manual creation provides maximum control over the sequence because the user decides exactly which postures are included, the order in which they appear, and the duration of each interval between them.

When one or more representative images of the task are available, they can be used to generate postures that are then incorporated into an animation. A posture can be entered manually from an image or captured using the artificial-intelligence posture-detection system (Section 14.3). This procedure is particularly useful when the task does not require a complete continuous reconstruction, but rather a representation of key phases, such as the beginning, an intermediate phase, and the end. Several images can be used to build a temporal sequence that is sufficiently representative for analysis.

Animations can also be generated in Ergoniza Intelligent by capturing multiple postures from a video of the activity. The system can load a video from the user's device, select the relevant time segment, and automatically record postures throughout the sequence using artificial intelligence. The video-capture interface allows users to choose between Fast AI and Accurate AI (Section 14.3) and, for Fast AI, set the sampling rate at which video postures will be recorded.

This procedure is appropriate for analyzing an actual activity performed by a person under natural working conditions. The user can select the video interval to be studied and automatically generate a posture sequence on the timeline.

When data from motion-capture systems are available, Ergoniza Intelligent can import them and convert them into an animation of the biomechanical model. In particular, the system supports the import of Xsens MVN data in MVNX format. This option is appropriate when a detailed temporal description of movement is required, especially in instrumented studies or complex tasks. Section 14.4 describes the import of data files from motion-capture (MOCAP) systems in greater detail.