Simple AI integration for Industry 4.0

As part of Industry 4.0, industrial companies aim to achieve the most efficient, flexible and networked manufacturing possible – and are increasingly using artificial intelligence (AI) to do so. Using AI, however, requires high-quality and ideally dynamic machine data.

In Industry 4.0, there are three challenges that must be solved to do so:

  1. At a typical manufacturing company, the process and plant levels are often characterized by a classic multi-vendor landscape – that is, the machines come from different manufacturers.
  2. The data of such manufacturing systems is heterogeneous and originates from components that have different technology standards, not the least due to their different ages.
  3. What is more, industrial systems have different communication interfaces and protocols. They generally generate and store data statically, which is then prepared manually by data scientists.

When artificial intelligence is integrated in these processes, this integration is often performed manually as well, due to the challenges described above.

Modular AI services for service-oriented architectures

The REMORA platform provides a remedy for such situations by simplifying AI development for Industry 4.0 and integrating AI services automatically. In the joint project “REMORA – Multi-Stage Automated Continuous Delivery for AI-based Software & Services Development in Industry 4.0”, the Fraunhofer Institute for Cognitive Systems IKS has teamed up with additional partners to work on the simple integration of AI services in Industry 4.0. Its goal is to simplify the integration of AI for the real-time analysis of machine data and to develop tools for high-quality, dynamic machine data. To do so, the project uses modular, flexible AI services within service-oriented architectures that can be easily exchanged and updated.

REMORA: Open, interoperable technology platform

The project is based on REMORA, an open, interoperable technology platform. Modular software tools called AI tech stacks enable the automated integration of dynamic analytics models, increasing the efficiency of AI applications in industry. The project partners are pursuing multiple goals with this platform:

  • A high level of automation for preparing data for the AI modeling process
  • A continuous, flexible supply of data
  • Meta-learning and automated data preparation across system boundaries
  • Integration of expert knowledge in AI models
  • Continuous further development of models

The framework supports and optimizes the AI life cycle

Within this project, Fraunhofer IKS is focusing its efforts on a framework for implementing an automated, continuous, dynamic data life cycle that consists of processes such as data acquisition, data processing, model training, data analysis and similar. The framework provides a library of different functions for AI services to enable simple integration of AI algorithms in Industry 4.0 applications.

In addition, the framework – which is compatible with existing Industry 4.0 standards and proprietary solutions – offers a concept for a comprehensive AI life cycle that focuses on three aspects:

  • Support for data scientists: The framework enables standardized access to data and metadata independently of the technology used. It also provides an overview of the manufacturing system topology, the collected data and its interrelationships, making it easier to understand the context of the data and develop new AI services.
  • Flexible integration of artificial intelligence: The framework enables flexible integration of AI services, for example, directly in the component or in the cloud.
  • Automation of AI processes: The framework automates AI processes, for example, through automated retraining and model deployment, which improve data analytics automatically and dynamically.

Integration of AI services through REMORA

Are you interested in using AI services at your company? The framework for optimizing the data and AI life cycle can be used in the entire automation sector: in manufacturing, mechanical engineering and robotics. Fraunhofer IKS will be happy to support you with the integration of artificial intelligence in Industry 4.0 applications. To obtain more information, get in touch with one of the contacts listed here.

This work is part of the project “REMORA – Multi-Stage Automated Continuous Delivery for AI-based Software & Services Development in Industry 4.0”, which is funded by the Bavarian Ministry of Economic Affairs, Regional Development and Energy and supported by Bayern Innovativ – Bayerische Gesellschaft für Innovation und Wissenstransfer mbH.

 

Industry 4.0

The term Industry 4.0 refers to the digitalization of production and manufacturing. In order to safely exploit the corresponding advantages, connected and automated production systems must function first and foremost in a dependable manner.