AI-based maintenance eases troubleshooting
How can user companies simplify the maintenance procedures for complex, connected machines to react as quickly as possible to failures, minimize downtime and reduce maintenance effort? This is the question the OWL scientists try to answer. Though their R&D project ADIMA (Adaptive Assistance System for the Maintenance of Intelligent Machines and Equipment) in practice is focusing on the machines in a large industrial-scale laundry, their solution can in principle be transferred to other types of equipment. Their goal is enabling the machines to automatically detect and identify any technical irregularities and displaying related maintenance and troubleshooting information by means of mobile devices, data goggles or projections. Thus, customers can react faster to potential failures, significantly increase the machine’s availability and, last but not least, reduce the efforts to deploy maintenance staff on site.
Within the scope of the project, the scientists will develop an assistant system that autonomously generates maintenance information based on Machine Learning algorithms, using machine data acquired locally, and visualize this information in a way that the troubleshooting routines can be performed by local technical staff even if this staff does not have machine-specific knowledge. “We want to develop ways to simplify maintenance and to perform focused error tracing at machinery with multimodal human-machine interfaces with the goal of reducing downtime and at the same time enable better working comfort for the operator through intelligent assistant systems,” explained project manager professor Carsten Röcker. Computer-based assistant systems, the Lemgo scientists are convinced, are the key to make the increasing complexity of machinery manageable for humans.
The developments arising from the project will be tested continuously for practicality. Therefore, a demonstrator will be implemented which will be connected to a real-world machine installation. The system to be developed will be based on the Internet of Things technology which means that data from very different sources are collected at the appropriate quality and then be processed and transformed according to the knowledge needed.
The business driver behind the project: If a machine vendor can reduce the maintenance effort and staff deployment, he can focus on the more profitable business with spare parts. First commercial deployments of the system are expected for 2019; besides the OWL University of Applied Sciences, two machine manufacturing companies are participating in the project. The German Federal Ministry for Education and Research is co-founding the project with an amount of € 600.000.
More information: www.init-owl.de
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :