Project Information

Project duration: 12/2017-11/2020

Project manager: Fraunhofer IVV

Fincancing: BMBF

Project partner: TU Dresden - Professur für Ingenieurpsychologie und angewandte Kognitionsforschung, Fraunhofer IVV Verarbeitungsmaschinen und
Entwicklung Verpackungstechnik, Theegarten-Pactec GmbH & Co. KG, Elco Industrie Automation GmbH, August Storck KG,  JR Die Schokoladenfabrik GmbH, TU Dresden - Professur für Prozessleittechnik

Project description

Processing machines produce mass consumer goods for daily consumption, where technical disturbances are unavoidable. The disturbance diagnosis is complicated by the complex plant structure, uneven force and torque progressions at high velocities and hardly modeled processing behavior due to networked, often unknown influencing parameters. In addition, different products need to be manufactured on the same machine which will be reinforced by the individualization in industry 4.0.

The goal of the joint project is the development of an adaptive, dialogue-bases assistance system (AS) for case-based closing for cooperative diagnosis and elimination of malfunctions in processing machines. To combine the potential of human and AS, the AS must (1) put the operator in a position to communicate the relevant information and (2) they must be able to interpret the communicated information on threir own to limit the range of the problem and to derive the requirements for the subsequent design of the dialogue. To enable this, the AS shall ask the operator’s current observations, their experience and conclusions based on it and relate them to the plant condition measured by sensors. This becomes steadily more specific in the course of the dialogue and finally leads to a diagnosis for the cause of the fault, on the basis which the operator can derive decisions and solutions strategies. The AS deliberately does not intervene in the system, instead its function is to create the basis for these decisions and interventions together with the operator. Over time, the AS learns from the dialogue-episodes, sets them in relation to the measured plant condition and develops its abilities to detect and communicate possible causes of malfunctions.