Computer Vision Optimization of Trench Shoring

Python
Computer Vision
Automation
API

Summary: Generating a bid off of an engineering drawing is a days to weeks long process for contractors. Our software was born of necessity to reduce this lead time. The overarching goal of this project is to generate the most efficient trench shoring configurations for a given excavation/new install of underground utilities given data collected from the engineering drawings using computer vision. Trench shoring is the act of reinforcing the walls of a trench to prevent collapse. From these configurations, we will use data collected from engineers and contractors in the underground industry to generate an estimated expense and time needed for an underground utility install using conventional shoring methods and an install that also incorporates the ShoreRight trench shoring system. The final results are then calculated to provide the user with a trench shoring configuration for the entire trench as well as the back fill for the entire trench. Workflow Explained: Our software begins with the engineering drawing of a single excavation site. From this engineering drawing we can use computer vision to extract almost all of the information we need to generate a trench shoring configuration. Once our software is running, the user will immediately be prompted with the values extracted by the computer vision from the engineering drawing. The user will be able to review and edit these values if there are any discrepancies. The user will then be presented with a simulation of the trench pictured in the engineering drawing. To continue, the user must click across the graph at the ground elevation. This is made simpler by following the dotted line in the underlying engineering drawing. Now the user can exit the simulation window and the software will use this user-input to generate the final shoring configuration. Once the simulation window is exited, a new window opens showing the ground level & trench floor, the location of all crossing utilities in the trench, and optimal placement of shoring. First Image: Provides an overview of the purpose and value of the project as well as a general overview of the implementation. This can also be downloaded in the artifacts section. Second Image: The raw data corresponding to the prior picture. This raw data has the start and end coordinate, size of shoring, type of shoring, and the number of shoring devices needed to be stacked on each other. Third Image: Example of an engineering drawing that can be used as the input file to our software. Fourth Image: Ability for user to edit the raw computer vision output is shown. Fifth Image: Provides the final shoring configuration output. Picture two creates a horizontal view of the specified trench and the optimal location of the shoring. Sixth Image: Zoomed in view of the trench shoring configuration, shows how the different trench shoring boxes fit around the crossing utilities (red dots) This project was proposed by team member, Sean Cameron, as a part of his broader work. The technical details are under strict NDA. For more information, please contact him directly at camerons@oregonstate.edu or see his associated work at https://shoreright.com/about/. The project mentor is Joe Louis, PhD, Civil Engineering, OSU. Special thanks to the collaborative efforts of Michael Carr - Branch Manager Murray Smith and Associates, Brandon Farmer - Project Manager Tapani Construction, Casey Lipe - Superintendent Tapani Construction and Tanner Whickwire - Project Manager Pacific Excavation

0 Lifts 

Artifacts

Name Description
Project Overview This artifact describes the purpose and value of the project as well as a general overview of the implementation.   Download
Work Flow Video In this artifact, we go through the process of running our software from start to finish.   Link