Work performance evaluation of heavy-duty mobile machines (HDMMs)

The construction industry is crucial for economic growth, but its productivity has not improved much despite its importance. Heavy-duty mobile machines (HDMMs), particularly excavators, play a central role in construction projects, with their productivity directly impacting projects' productivity and costs. This dissertation aims to tackle several challenges regarding the automatic productivity estimation of an excavator in earth-moving operations, such as loading, trenching, and grading.

In the beginning, the significance of the construction industry and the critical role of HDMMs within it are discussed. It highlights the challenges faced by the industry, including low productivity growth and outdated practices, emphasizing the need for automated productivity estimation and progress monitoring. Then, an excavator is introduced as the main application in the research study. In the next phase, existing research studies for the productivity estimation of HDMMs are thoroughly explored to identify research gaps and to design multiple research questions that drive the dissertation's focus.

Capturing motion information using inertial measurement units (IMUs) holds promise for recognizing activities and automatically estimating cycle time and productivity. ... mehrAlso, the importance of analysis of working conditions and estimating theoretical cycle time and productivity is stated. In addition, 3D sensors and building information modeling (BIM) can be integrated to enhance the productivity estimation and progress monitoring of an excavator in quality-centered tasks, such as grading and trenching operations.

First, an activity recognition method is proposed to identify the excavator working cycle using supervised classification methods and motion information, such as angular velocities and joint angles, obtained from four IMUs attached to moving parts of an excavator, including the swing body, boom, arm, and bucket. Human operators perform tasks using a medium-rated excavator under different working conditions, such as different types of material, swing angle, digging depth, and weather conditions to collect a dataset. The proposed method can effectively recognize the working cycles of an excavator. Task recognition can aid management teams in monitoring productivity and progress, optimizing resource allocation, and scheduling. Using the results of the task recognition algorithm, productivity can be calculated based on task-specific metrics.

Next, an approach is designed to automatically determine the productivity and operational effectiveness of an excavator in the loading operation. Firstly, an algorithm is proposed to recognize the excavator's sub-tasks using supervised learning and motion data obtained from IMUs. Then, a method is presented to estimate the actual cycle time based on the sequence of activities detected using the trained classification model. The actual cycle time cannot solely reveal the machine's performance since operating conditions can significantly influence the cycle time. Therefore, a reference is required to analyze the actual cycle time. Secondly, the theoretical cycle time of an excavator is automatically estimated based on the operating conditions, such as swing angle and digging depth. Thirdly, the relative cycle time is obtained by dividing the theoretical cycle time by the actual cycle time. The relative cycle time index can effectively monitor the performance of an excavator in loading operations and can be useful for worksite managers to monitor the performance of each machine in worksites.

In the next step, a technique is proposed to estimate the excavator’s actual productivity in trenching and grading operations. In these tasks, the quantity of material moved is not significant; precision within specified tolerances is the key focus. The productivity definitions for trenching and grading operations are the trench's length per unit of time and graded area per unit of time, respectively. In the method, a height map from working areas is constructed. Also, BIM is utilized to acquire information regarding the target model and required accuracy. The productivity is estimated using the map comparison between the working areas and the desired model. The method can effectively estimate productivity and monitor the progress of these operations. The obtained information can guide managers to track the productivity of each individual machine and modify planning and time-scheduling.

This dissertation employs advanced technologies, such as IMUs, machine learning techniques, elevation terrain mapping algorithms, and BIM. It aims to streamline productivity estimation and progress monitoring for excavators, ultimately contributing to more efficient and successful construction projects. It underscores the potential for future research to enhance these methodologies, expand their applicability to other HDMMs and tasks, and address remaining challenges to propel the construction industry towards greater productivity and sustainability.