Today’s business environment sustains mainly those companies committed to a zero-defect policy. Therefore, the prompt detection of rare quality events has become an important issue and create an opportunity for manufacturing companies to move quality standards forward. Machine learning as a tool has been applied in many fields including quality control, which is the task of assuring that all products produced reached a certain standard. On the other hand, industrial robots have been widely applied to the automation of industrial production processes.
This project integrates machine learning with pattern recognition strategy and a robotic system to realize an intelligent part inspection system, in which the main goal is the detection of part defects and different configurations. The project strives to build a successful decision-making procedure and successful implementation of machine learning in part quality control. Overall, this project has demonstrated the potential to vastly improve the efficiency of operations while reducing defects, helping to cut quality, costs and improve customer satisfaction
Technology Features, Specifications and Advantages
The automated system in the project integrates several technology features including the following:
- Machine learning - A machine learning algorithm has been developed for part inspection based on SAS machine learning engine. The solution is formulated as a binary classification problem of machining learning. SAS machine learning model is applied for the model training and scoring.
- Microprocessor - Raspberry PI is used as the operating system for its’ advantages in being small in size, low cost, and ease of installation. It controls the whole system processes including taking pictures, connecting to SAS machine learning engine, sending the command to Arduino system which further controls the robotic arm.
- Robot control - A robotic arm is controlled by Arduino board and program. The Arduino board receives the scoring result of the specific part.
- Software programming - A mixed programming systems are applied and integrated with the system, including Python, networking, Arduino, etc. Python is used to develop the central system. Arduino is applied for the control of the robot.
The approaches applied in the project can be widely applied to manufacturing processes to boost the performance of traditional quality methods and potentially move quality standards forward. More specifically, it can be applied in the potential areas such as quality control and management, part feature detection, part configuration, and classification.
- Quality improvement – benefit from the on-line quality control
- Productivity improvement – benefit from the automation system
- Manpower cost reduction – benefit from the automation system