KIKA-IPK
AI cognition-supporting assistance system for in-process control in manufacturing
PROJECT DESCRIPTION
AI-supported image processing and assistance functions
OBJECTIVES
The objective is to develop an AI cognition-supported assistance system for in-process control (KIKA-IPK), which enables a more resource-efficient process and material configuration through self-learning image feature correlations with process properties. Here, the experience knowledge of the machine operator for the connection of visual quality features on the one hand and process properties on the other hand is modeled by machine learning methods. As a project result, an assistance system is targeted that enables a more resource-efficient configuration of process parameters by mapping visual quality features of the product and its process variables in an AI model during manufacturing.
INNOVATION & METHODOLOGY
Within the scope of the R&D project, methods are being developed that make it possible to draw conclusions from visual product characteristics about process parameters that can hardly be measured. This makes it possible to control the process in such a way that quality deviations are compensated for during production and efficiently adapted to new product characteristics. For this purpose, the image, process and material data streams as well as user feedback during production are analyzed via the interface of the "AI cognition-supporting assistance system" (KIKA) and the results are comprehensibly transmitted to the actors as well as the machine control in real time. The AI services are integrated into two scenarios for additive manufacturing, 3D metal printing with steel and personalized medicine printing, and the resource efficiency potential is demonstrated in industrial applications.
OUR CONTRIBUTION
The technological goal of Gestalt Robotics is primarily the expansion of the technology portfolio to include active learning services with the integration of user feedback. In this way, a technological bridge is created between existing application areas of AI-supported image processing and novel application areas in the field of intelligent assistance systems. In addition, the industrial application of explorative learning methods, e.g. reinforcement learning, can be piloted within the project framework. The concrete result is a recommendation system to support the machine operator in the visual product characterization by ML methods during production and to output corresponding measures for the compensation of quality deviations as specific instructions for the machine operator and the control system.
© Technische Universität Berlin
Key facts
Assistance in Manufacturing
Real-time machine and process control
Cognition support based on experience and knowledge
Resource-efficient process and material configuration
Machine Learning
Self-learning image feature correlations with process features
AI revision based on user feedback
Synthetic data and artificial generation of real-time process knowledge
Service Cloud and Machine Connectivity
AI cloud services and platform
Open machine interfaces
Real-time connectivity and WebUI
PartnerS & CONSORTIUM
BioFluidix GmbH
Demonstration of in-process control for real-time compensation of quality deviations in the printing process and performance evaluation
DiHeSys GmbH
Demonstration of the KIKA-IPK assistance system and potential evaluation in personalized medicine
GEFERTEC GmbH
Demonstrate in-process control, develop strategies to adapt to new product features, and evaluate potential for metal 3D printing
PSI Metals GmbH
Interface development, performance evaluation of AI cloud services, and potential evaluation for metal fabrication
RELIMETRICS GmbH
Implementation of (ML) methods for product feature-based process control by artificial generation of real-time process knowledge from image information
TU Berlin
Development of a human-AI hybrid (ML) method for self-learning image feature correlation with process features for autonomous in-process control and process control
Funding
Federal Ministry of Education and Research
Funding Initiative
Lernende Produktionstechnik – Einsatz künstlicher Intelligenz (KI) in der Produktion (ProLern)
Program
The Future of Value Creation – Research on Production, Services and Work
Duration
11.2021 – 31.10.2024
Project Management Agency
PTKA Projektträger Karlsruhe
Project Poster
Project poster (German) as printable PDF
This project is/was financed with funding provided by the Federal Ministry of Education and Research under the “The Future of Value Creation – Research on Production, Services and Work” program in the call "Lernende Produktionstechnik – Einsatz künstlicher Intelligenz (KI) in der Produktion (ProLern)" and managed by the Project Management Agency Karlsruhe (PTKA). The author is responsible for the content of this publication.