Processing equipment cleaned flexibly and with agility thanks to a new device
 |
| Mobile Cleaning Device in two versions: self-driven and on-conveyor-belt (Photo: © Fraunhofer IVV Dresden) |
Fraunhofer IVV Dresden has designed an innovative mobile device in order to assist in cleaning the processing equipment. The showcase of the Mobile Cleaning Device (MCD) at VDMA booth (J38 in Hall 5) will highlight the advantages of combining the reliability of conventional automated cleaning systems and the flexibility of cleaning by hand. The MCD system is equipped with optical sensors enabling the detection of dirt and adaptability. Which zones require cleaning? How long has it been since the last cleaning? Was the cleaning successful? The future in-line systems will make it possible to answer all these questions.
The virtual twin of the MCD system displayed at the booth includes an adaptive model of the cleaning process. Its combination with the concepts of cognitive control and a system of sensors for dirt detection results in the premiered adaptive cleaning. One should understand this term as cleaning adapted to the hygienic condition of the equipment. The system versatility is once again confirmed by the drive concept. Movements between machine units can be enabled by a dedicated drive unit or by preexistent systems of transportation, e.g. conveyor belts. Contrary to conventional cleaning systems, the MCD does not cover one particular machine but can be used flexibly to clean a certain number of machines. Nozzles for spray and foam cleaning are available as individually driven units. In addition to cleaning entire machines, directed cleaning of their parts is possible.
Self-learning assistance system designed for machine efficiency improvement
Fraunhofer IVV Dresden enthusiastically creates self-learning assistance systems for processing equipment, the first concepts for which are about to be displayed at Interpack 2017. One should admit that even the state-of-the-art production lines are predisposed to brief stoppages every five minutes; that was the exact pivot for developing these systems. Today's processes and equipment are more and more complex. Oftentimes, production line operators cannot remedy faults they encounter at their very source; they only manage to minimize the immediate effects. And even the most advanced sensor systems aren't able to avoid 100 percent of faults, e.g. those due to inconsistent product quality.
The expertise of skilled, highly qualified machine operators is still the key knowledge source for rectifying equipment faults. This valuable knowledge has to be taught to less skilled staff. That is the reason for Fraunhofer IVV Dresden to follow a number of concepts providing machine operators with data on fault elimination relevant for the dominant conditions. The concepts in question have finally been integrated in SAM, the Self-learning Assistance system for Machines, which assists operators in rectifying faults thanks to a quasi-navigation system. The system relies on identification of anomalies in sensor signal patterns, based on machine learning and data mining practices. A fully cooperative dialog system is on its way. It will make the assistance system learn directly from operator input and propose, in cooperation with the latter, an issue-solving strategy. The SAM itself will not actively engage in the process.
