Entering a confined space to perform work is dangerous.Despite numerous requirements established by OSHA under the Department of Labor, injuries and fatalities in confined spaces occur every year.In 2015, according to data compiled by the Bureau of Labor Statistics, there were 136 fatalities associated with confined space entry.
OSHA requires training for personnel entering confined spaces which must include, at a minimum, the following (https://www.osha.gov/ ):
The safest practice, however, is to avoid entering a confined space.Advancements in robotic systems are making this possible.For example, technicians and engineers often enter confined spaces for the purpose of performing inspections.These inspections range from visual detection of damage to the use of sophisticated NDT tools to identify damage.These inspections are necessary to determine the fitness-for-service of the assets under inspection.
Recently, Seikowave and GE Inspection robotics combined their tools and engineering talent to use robotic crawlers and 3D imaging to replace humans for the inspection and assessment of damage in confined spaces.With support from Chevron and Texas A&M, we were able to deploy the combination of the GE FAST RVI and the 3DSL Rhino inside several confined spaces.On a gray, overcast day, we conducted testing in the rail tank car shown below.
This tank car is typical of the more than 350,000 rail tank cars in operation today.In the US, the Department of Transportation has strict inspection requirements for the assessment of rail tank cars.Among many parameters that operators are required to assess, the wall thickness of the rail tank car is critical.Currently, identification of areas of damage inside the rail tank car are primarily performed visually.Once the technician or engineer has identified the area of damage, further inspection is performed to determine the wall thickness.Today, this work is done by personnel entering the rail tank car to perform the work.
The FAST RVI robotic crawler from GE is shown below with a 2D inspection camera and a 3DSL Rhino.We tested the following workflow:
Once fully assembled, the inspection robot was placed into the rail tank car.The image below shows the tool at approximately the 9 O’Clock position acquiring 3D data.
Below is an image of 3D data of rail tank car damage overlaid with the texture image to provide a photo-realistic view of the area that was inspected.This data was taken at the 6 O’Clock position.The damage is barely visible and was, in fact, not noticed by the engineers and technicians performing the work.However, once the data was analyzed for metal loss, the extent of the damage could be determined.The damage in this area ranged from 0.004 inch to 0.018 inches with an error of +/- 0.002 inches.The diameter of the defects were small: on the order of 0.040 inches or less.
The use of a robotic crawler combined with 3D imaging enable inspection in a confined space without the need for entry into the confined space.The inspection identified metal loss that was not noticed by either personnel viewing the 2D image or directly looking at the area of damage.We believe that the uniform color of the steel, the low lighting conditions, and the diameter of the defects made it difficult to identify the area as damaged using conventional visual techniques.The 3D imaging tool used in this study, the 3DSL Rhino, is not affected by the low light conditions or uniform surface reflection of the inside of the rail tank car.We believe that this approach will improve the quality of inspections, thereby improving rail safety, and reduce the need for personnel to enter confined spaces.
For more information, contact Seikowave.