Predicting Optimum Parameters for Straightening Machines for Manufacturing Oil Drilling Pipes
THE BACKGROUND
Our client is a leading global producer of premium seamless pipes for the oil and gas industry. Seamless pipe production involves multiple stages of continuous and discrete processing. The melting stage generates billets which are solid round tubes. The rolling process takes these tubes and goes through a sequential process of piercing, milling, heat treatment, and stretch reduction to generate ‘mother’ tubes. These tubes are then cut into pieces and sent to the finishing facility. The lead end (or the front of the tube) suffers from a defect called ‘fish mouth’, which impacts ovality. This leads to threading defects towards the end of the process known as Black Threads. Black Threads leads to wastage and additional costs in re-processing pipes.
As part of transforming its manufacturing process by formulating Industry 4.0 initiatives, our client wanted to minimize thread cutback (black threads) through pipe and straightener control and have straightening measurement values within acceptable range. Soothsayer collaborated to build an Artificial Intelligence-driven straightener model that predicts optimal values for parameters that define the straightening process, given an incoming straightness value to maximize the probability of getting acceptable output straightness.