Drip Pipe Drilling Quality Check with AI – Case Study

This case study explains how our client has benefited from a data-driven approach to drip pipe drilling. By using artificial intelligence (AI) they have been able to improve quality while reducing costs and improving safety.
groove and non-groove detection - drip pipe drilling quality checking

Our client has been a market leader in irrigation equipment and supplies for years. They have a large selection of products that are sold in many countries. Their manufacturing process is automated to reduce human error, increase efficiency, and save time. An essential step in making a drip line pipe is to drill holes into a pipe as per specifications. Accuracy is essential in this step, or material will be wasted.

Opportunity

Our client was looking for a Computer Vision solution provider to help modernize the drilling and targeting process. The company aimed to increase accuracy of its drilling and reduce waste to boost profits. 

When they reached Visionify.ai, their proposal was centered around developing the following.

  1. Develop a laser triggering solution for drilling holes on drip line pipe by using computer vision.
  2. Computer Vision solution to check quality of holes to ensure they meet specifications.
  3. A real-time feedback solution that monitors pipes’ position and adjusts it if necessary.
  4. Real-time notification solution when stack light will indicate an out-of-bounds error.
  5. Business analytics to provide a bird’s-eye view of process performance, drilling, detection, statistics and quality metrics that matter.

Challenges to be solved

  1. To check the quality of the drip pipe hole drilling process with different pipe measurements that include tube diameters, wall thickness, emitter types, and lengths. 
  2. To check the Grooved and Non-Grooved Areas of the pipe

Solutions provided by Visionify.ai

Targeting solution

To solve the issue we built a computer vision-based targeting solution to improve accuracy of the drip pipe drilling process. That works for up to 1000ft/minute pipeline speed within target tolerance limit. This solution can be integrated and worked with the existing laser drilling process.

Quality Check Solution

The quality-check solution uses Computer Vision technology to monitor hole drilling processes. Gives feedback to targeting and laser systems in real-time if the target moves in either direction, up or down. Both solutions provide a live stream of data, deviation graphs, and other vital statistics on the deployed monitors. Our solution also detected grooved and non-grooved areas of the pipe as shown in the images.

What Next?

Are you facing any production issues while manufacturing irrigation supplies? Call us. We will help you resolve those issues in a few weeks and get a demo of our factory vision solutions and how we can integrate our solution with your production lines.

Results

ROI (Region of Intrest) the section in the middle of the blue lines is our Region of Intrest. Detection of Accuracy for grooves and non-grooves area in irrigation pipe is more than 85%.). Average Precision for holes is 62.401 and for grooves is 61.383.

Follow Visionify.ai on LinkedIn & Twitter

detecting grooved and non grooved area
mask rcnn
fast rcnn