“I’m 100% confident this is the exact location to dig….I think.” How often does this happen when making the decision of where to excavate and uncover pipe and point repair defects? In many cases, it may require multiple locations and excavations before the correct location is discovered. This can dramatically increase the cost of point repairs due to varying pavement types, obstructions and or other utility replacement or relocations. Typically, the location to excavate is primarily taken from CCTV footages and may or may not be as accurate as you think.
Pay out. I am not talking about your local casino. Rather, I am talking about the pay out distance from the manhole that serves as the primary means of locating defects within your pipes. What is the actual location of pipe defects and features as recorded in CCTV, and is it important for that measured payout distance to be accurate? Said differently: how often are you digging in the wrong place?
I have noticed that much of our work focuses on defects and error tolerance as it relates to defect location. If a robot can locally detect a defect, classify its severity to sub centimeter accuracy, can the inspection robot tell you where to dig? More importantly, how close is close enough?
The first problem we deal with is related to the camera. Best practices in PACP and WRc inspection require pipe defects and features to be coded with cameras perpendicular to the pipe centerline, and with strict cable controls to ensure accurate location. That means, the camera is just under, just over, or just next to the defect. But zoom cameras are just so tempting! Even experienced operators like to position their camera in a static position and zoom forward to capture the defect. The result is that we have the payout from the manhole measured accurately, while the actual sewer problem may be many feet away. In large pipes, we have seen CCTV tapes with defects and pipe features coded 20-30 feet in advance of their true location.
But the issue doesn’t stop there. Even camera operators who manage their inspections strictly find that non-zoom issues, such as cable stretch, pipe geometry, and hydraulic action, can cause location errors that are measured in 10’s of feet. And finally, we have the issue of, where does the robot go after being lowered down into the manhole?
RedZone is working on this problem and looking to provide reports that specify the distance from the manhole within a certain error tolerance. Let me tell you how we hope to do it. Unfortunately, I have to start with a lecture.
We all have to understand the value and downside of zoom cameras. That zoom can enable your robot to sit in one place and see all the issues. But once the defect is located, operators have to guide the robot to it and take a perpendicular picture. Not only is that a best practice, but it enables us to use other technologies that tells you where to dig. So let’s learn about that other technology.
We’re all familiar with GPS. With GPS in a car, we can find our coordinates, then drive our car a hundred feet in a known direction, say North, into a tunnel (where the GPS doesn’t work), add a hundred feet to the GPS reading, and have a pretty accurate picture of where our car is. This same concept enables you to understand where to dig. However, the way we measure the distance from the GPS reading to the “tunnel” is pretty sophisticated, and is called IMU – Inertia Measurement Units.
IMU records the rates of acceleration in three perpendicular axes and uses gyroscopes to detect subtle changes in roll, pitch and yaw. Acceleration measurements yield the distance moved along the three axes while the gyroscope provides the rotations with respect to the same three axes. Used in a sewer robot, an IMU can continuously track the motion of the robot as it is inspecting the pipe and this data can then be used to estimate the x, y and z location of the robot in the pipe. Since it is not possible to use GPS in a sewer pipe, getting the x, y, z locations using an IMU is the next best thing.
Manholes help us take the IMU locations obtained from within the pipe to the next level. At the start of an inspection run, we get the GPS location of the manhole that the robot is being deployed into – since GPS is available at that above-ground point but not anywhere else inside the sewer line. Starting from this GPS location, each of the below-ground points estimated using the IMU can be translated to world co-ordinates. This is done repeatedly with the IMU data on a sewer robot to accurately estimate where the pipe is during the whole inspection run. This pipe location can then be plotted into a plan and profile view of the pipe centerline. This is the Holy Grail that provides exact locations of critical defects, important pipe construction features, and enables precision geo-location of our underground assets.
With an IMU, we are able to solve two of the three problems that underground camera inspection creates: where did the robot actually go, and how was the robot’s travels affected by the terrain, by the distorted payouts, by the hydraulic action? This is because we track the robot in three dimensions. However, we can’t use it to track operators who use their zoom lenses instead of coding with cameras perpendicular to the pipe centerline.
Much work is to be done on this area before a comprehensive solution to defect location and pipe positioning is standard practice in the industry and we need your help. Review inspection tapes from your most critical assets and check for proper coding and cable management practices. I think you will be amazed by what you find.