Corporations are stepping up their efforts in transformation to smart manufacturing to cope with the advent of the age of AI. An "AI Experimental Platform" which was built in collaboration with Hubei University of Technology in China was born out of this new digital trend. Instead of adopting the conventional industry-academic cooperation model, Delta aims to cultivate new generation of AI talents to meet industry needs.
Consisting of more than 20,000 teaching staff and students, and focusing on engineering education, Hubei University of Technology in Wuhan is one of the key universities to carry out the "Excellent Engineer Education and Training Program". The "AI Experimental Platform" project establishes the production line on campus and aims to promote the core concept of "intelligence". The platform not only enables both teachers and students to plan related courses according to actual needs of teaching and learning, but also can repeat practices at any time. Both student needs for experimental learning and teacher needs for R&D scenarios can be met; the contents taught by the teachers can be internalized more directly and immediately in the form of practical student skills that connect seamlessly with the needs of enterprises. This new industry-academic cooperation project, scheduled to be officially launched in the new semester of 2021, is expected to cultivate more industry’s talents and inject new blood to the industry.
The solutions deployed by "AI Experimental Platform" include:
Transformable intelligent machines (TIM): TIM meets the needs of "small in volume but diverse in variety", the standardized cabinets of TIM developed by Delta Research Institute can connect in series with various robotic arms and functional application modules. Depending on the need, machine functions can be disassembled and assembled whenever necessary.
Automatic defect classification: The image data of AOI machines and utilizing visual computing and image analysis to perform deep learning and construct models, the automatic defect classification software introduced by Delta Research Institute can improve identification of nonconforming product and correct classification rate.
Predictive maintenance: Monitoring data from equipment, system and sensors are integrated with AI algorithms to establish parametric models, the predictive maintenance solution can realize status monitoring and anomaly detection. Some of its modules are open to teachers and students to conduct secondary development and experiments.
PC-based control system: Software programs in the PC-based control system can be controlled by dragging & dropping charts, which enables faster execution of machine and equipment software based on any production changes. It allows users to control the TIM and edit production processes.
Composed of intelligent TIM hardware and intelligent software - automatic defect classification, predictive maintenance, PC-based control system and more, Delta’s "AI Experimental Platform" reproduces actual industrial application scenarios at universities.