Heavy equipment accounts for a large portion of construction site accidents. The construction schools, therefore, use actual equipment and simulators to train equipment operators to work safely. Existing training with the actual excavation equipment helps trainees to get sensitized to delicate details about machine operations. The feedback process during on-machine training seems, however, suboptimal. This is mainly because feedbacks are currently provided by means of visual inspection performed by the instructors, and mainly communicated verbally. This practice has several problems:
Attention focus and span: During the stretch of the on-site lessons the instructors have to divide their attention – for observing, analysing, memorizing and feedback - across several students at the same time.
Continuity: Direct immediate feedback would break the training sequence of the student;
Transience: Post-hoc feedback creates the problem of remembering and reconstructing of the sequence for both student and instructor;
Perspective and factuality: Students experience the lessons as actor/operator from an insiders perspective. The instructors observe from an outsiders perspective. This can create student-instructor misunderstanding regarding the movements as actually preformed;
Proximity: Instructors need to always maintain a safety distance due to the safety risks involved in moving near the equipment. Furthermore, the large sizes of the construction equipment may block trainers views;
In brief, instructors may experience difficulty in comprehensively observing, memorizing and analyzing trainees’ moves. Consequently, many important operational details/manoeuvres can be overlooked by the instructors when training outside. Further, since the actual on-equipment training is relatively expensive, it is, therefore, desired to develop more efficient solutions that get more out of the trainees’ interactions with the equipment.
Accordingly, it is very favorable to devise a new solution that uses data collected from various sensors on the equipment to track and visualize the performances of trainees and, thus, support the instructors in providing feedbacks to the trainees. Advanced visualization technologies allow for rapid reconstruction of scenes in the virtual environment, which can be used to support the instructors in better pinpointing the major and minor mistakes of the trainees. Another advantage is that the training performances can be archived virtually to support detailed progress monitoring and personalized training.
The aim of this design project is to:
“Develop a feedback support system that can assist the instructors of the heavy construction equipment to provide a more comprehensive guidance to the trainees”
The main outcomes are:
An easy-to-mount toolkit for tracking the motions of the heavy equipment on the training site;
A visualization-based feedback system that uses the sensory data to reconstruct a rendition of the trainees’ performance;
An algorithm to identify and demarcate the “points of attention” in the performance of the trainee for the instructors;
A guideline for the incorporation of the developed system in the exisitng educational process.
Deadline Application: 01-10-2017