Location
Hilton Hawaiian Village, Honolulu, Hawaii
Event Website
https://hicss.hawaii.edu/
Start Date
3-1-2024 12:00 AM
End Date
6-1-2024 12:00 AM
Description
This paper presents an approach that utilizes artificial intelligence techniques to identify autonomous machine behavior patterns. The context for investigation involves a fleet of prototype autonomous haulers as part of a Product Service System solution under development in the construction and mining industry. The approach involves using deep learning-based object detection and computer vision to understand how prototype machines operate in different situations. The trained model accurately predicts and tracks the loaded and unloaded machines and helps to identify the data patterns such as course deviations, machine failures, unexpected slowdowns, battery life, machine activity, number of cycles per charge, and speed. PSS solutions hinge on efficiently allocating resources to meet the required site-level output. Solution providers can make more informed decisions at the earlier stages of development by using the AI techniques outlined in the paper, considering asset management and reallocation of resources to account for unplanned stoppages or unexpected slowdowns. Understanding machine behavioral aspects in early-stage PSS development could enable more efficient and customized PSS solutions.
Recommended Citation
Aeddula, Omsri; Ruvald, Ryan; Wall, Johan; and Larsson, Tobias, "AI-Driven Comprehension of Autonomous Construction Equipment Behavior for Improved PSS Development" (2024). Hawaii International Conference on System Sciences 2024 (HICSS-57). 4.
https://aisel.aisnet.org/hicss-57/da/manufacturing/4
AI-Driven Comprehension of Autonomous Construction Equipment Behavior for Improved PSS Development
Hilton Hawaiian Village, Honolulu, Hawaii
This paper presents an approach that utilizes artificial intelligence techniques to identify autonomous machine behavior patterns. The context for investigation involves a fleet of prototype autonomous haulers as part of a Product Service System solution under development in the construction and mining industry. The approach involves using deep learning-based object detection and computer vision to understand how prototype machines operate in different situations. The trained model accurately predicts and tracks the loaded and unloaded machines and helps to identify the data patterns such as course deviations, machine failures, unexpected slowdowns, battery life, machine activity, number of cycles per charge, and speed. PSS solutions hinge on efficiently allocating resources to meet the required site-level output. Solution providers can make more informed decisions at the earlier stages of development by using the AI techniques outlined in the paper, considering asset management and reallocation of resources to account for unplanned stoppages or unexpected slowdowns. Understanding machine behavioral aspects in early-stage PSS development could enable more efficient and customized PSS solutions.
https://aisel.aisnet.org/hicss-57/da/manufacturing/4