The AI-PRISM project is oriented towards a human-centred AI-based solutions ecosystem targeted to manufacturing scenarios with tasks difficult to automate and where speed and versatility are essential. To facilitate the assessment of the performance, transferability, scalability, and large-scale deployment of AI-PRISM solutions, we will conduct research under real operational environments in five pilots involving key manufacturing sectors: furniture, food/beverage, built-in appliances, Electronics and one generic discrete manufacturing.
Regarding the built-in appliances manufacturing demonstrator, researchers from Teknopar (TEK), Cranfield University (CRAN) and Comau (COM) had the opportunity to visit Silverline (SIL) premises in Merzifon, Turkey. Moreover, other team members attended virtually and reviewed the envisaged process implemented in the pilot, such as ETRI and A&G from South Korea. The former has been the largest Korean government-funded research institute since 1976. The latter will contribute to integrating and validating the quality collaborative control workstations for end-of-line tests.
In the Silverline production plant, there are different production stages, such as sheet metal processes, powder coating, bonding of the glass pieces and assembly lines. At the end of each assembly line is a Quality Control Box to perform Final Quality Control and clean the product before the packing operation. This use case focuses on range hood production’s final control workstation, which application of AI-PRISM solutions will help increase the efficiency of resource use (with subsequent reduction in production costs and increase in production volume) and the safety of human operators.
The data collection experience in the control workstation
At the final control workstation, operators perform visual control, grounding test, functional test, cable packing, sticking of the packed cable, sticking the metal label inside the product, filter assembling, bagging the product process and putting barcode labels and documents near the products. Three operators take 30 seconds to perform them, and risks found are quality defects, delays, errors, and defective and missing deliveries. In addition, some impacts on these operators are stress and less job satisfaction, such as safety and health problems.
During their visit, the tasks analyzed by AI-PRISM researchers were the barcode reading and label put inside the good, visual inspection, buttons testing, Earth bonding testing, fan testing and lights testing. More concretely, researchers eye-tracked three processes: fan grounding, visual inspection and labelling. Through this data collection, they detected that within the same work cell some steps are being duplicated which might be increasing the workload for the operators. By integrating tasks into one process, operators can utilize their time for other tasks or specialize on other processes.
Challenges and next steps in the pilot
Currently, there are no robots or robotic frameworks, data models or sensors, communication infrastructure, or data platforms in the use case. Therefore, one significant challenge of this pilot will be the robot’s integration and testing capabilities. Especially visual control for the finished product, which decision will affect the successful result for our use case.
A new simulation platform that will enable the design and simulation of the final control of the range hood production line, including the interaction between existing agents, needs to be developed during the project. Simulation software (e.g., Gazebo, ZeroSim- Unity) will be examined for this procedure. In addition, robotic agents need to be integrated with the URDF model, a model for specifying the geometry and organization of robots in ROS.
After the visit, the AI-PRISM team will conduct questionnaires and interviews to identify the most unwanted operations to reduce cognitive load, support Silverline for solution identification (simulation of reachability/cycle time) and propose cameras for visual inspection in the pilot. The AI-PRISM team is excited to see where this project will take us. We will apply the same methods and technologies in the other industries that conform to our five pilot use cases, requiring high manual dexterity and expertise.
Do you want to learn more about AI-PRISM and the team behind the project? Subscribe to our newsletter and follow us on LinkedIn and Twitter, so you don’t miss a thing!