How We Used AI and Robotics to Scale Orthopedic Screw Production

When the COVID-19 pandemic hit, it turned industries upside down but particularly shook healthcare to its core. Among those struggling with the aftermath of the crisis was one of Techolution’s prominent life science clients - a medical device manufacturer specializing in orthopedic screws. Unsure, yet curious about the potential of Artificial Intelligence, they turned to us to solve their biggest problem - scaling the production of their orthopedic screws.

Enhancing Quality Control In Bone Screw Production

During our initial engagement, we visited their factory and interacted with floor managers to identify their primary challenges. We learned that after the COVID-19 pandemic, there was a significant labor shortage. Our client struggled to hire and retain enough workers to perform the meticulous quality control process required to produce high-quality bone screws.

We began uncovering the root cause of the problem instead of proposing ready-made solutions. As we followed our practice of ‘building with customers’ we understood that the current process of quality control involved using vernier calipers to individually measure each screw. This manual process required training, took approximately 2 minutes per screw, and was prone to human error. Moreover, the company couldn’t raise prices to hire more quality control workers due to market constraints, leading to an inability to meet growing demand. That said, accelerating defect detection in these screws was a critical real-world problem that we knew AI could solve.

When It’s About Hair-Width Precision, There’s No Room For Error

Our client has a stringent margin of error of 0.01 mm or 1/10th of the width of a strand of hair which is not only an internal requirement but also mandated by the U.S. Food and Drug Administration (FDA). Their regulations required manufacturers to randomly sample and precisely measure 13 out of every 100 bone screws, documenting these inspections exhaustively.

Recognizing the severe consequences of using defective screws such as chronic pain or permanent disability it was crucial to speed up defect detection with utmost accuracy. Given our expertise in AI-powered image recognition, robotics, and edge device fabrication we proposed a cutting-edge solution to address the manufacturer's challenges. We devised a robotic arm equipped with a high-definition camera capable of measuring and sorting orthopedic screws as per the FDA standards.

Meet Alice! She Boosts Profits And Serves Drinks

We named this system ‘Alice’ which works by taking a picture of a batch of screws and using computer vision AI, detects where the screws are in the image. Alice then picks up each screw, positions it in front of the camera, and converts pixels into millimeters to detect any defects. Screws within the acceptable threshold are placed into the ‘pass’ bin, while the defective ones are discarded into the ‘fail’ bin.

Our AI technology transformed the manufacturer’s quality control, cutting costs, time, and manual effort while significantly reducing errors. The turnaround time for defect detection dropped from 2 minutes per screw to under 10 seconds. This efficiency allowed the company to scale their production of orthopedic screws to meet the growing demand and significantly boost their revenue and profitability. Additionally, the entire process was digitally recorded, enabling real-time creation of the FDA-mandated documentation which eliminated the tedious and labor-intensive paperwork.

While Alice proves that she can accelerate quality checks, she is also great at bartending. That’s right, Alice has a fun side to her as well. Her revolutionary use of built-in sensors and depth camera to take 3D pictures to open doors, grab and hand you drinks showcases her versatility and ability to coexist with humans.

Meet Alice, our favorite bartender