Summary
- Profile Type
- Technology offer
- POD Reference
- TOSI20230315015
- Term of Validity
- 14 April 2023 - 13 April 2025
- Company's Country
- Slovenia
- Type of partnership
- Commercial agreement with technical assistance
- Research and development cooperation agreement
- Investment agreement
- Targeted Countries
- All countries
Contact the EEN partner nearest to you for more information.
Find my local partner
General information
- Short Summary
- A group of innovators developed an innovative approach using advanced techniques inspired by nature and machine learning (ML) for predictive maintenance in industrial processes. It automatically finds optimal ML models for each case, saving time and money. It allows companies to monitor and optimize industrial processes and increase productivity and safety. The method is flexible and can be used for various cases depending on user parameters and data.
- Full Description
-
The innovation is designed for use in predictive maintenance in manufacturing plants, utilizing advanced methods inspired by nature and machine learning. The main feature of the method is its system, which can adapt to different instances of predictive maintenance and automatically finding the optimal technique. This approach can save time and financial resources in practice, as it reduces the need for manual steps in developing machine learning models. In the classical approach to using machine learning for a given problem, a model must be defined and constructed for each use case, a process that also requires expertise and knowledge in the field. Due to the complexity of the data and the decisions associated with model architecture, experts may subjectively choose a solution technique. With the proposed method, this process is automated and left to a computer algorithm, which iteratively finds the objectively optimal solution based on various evaluation metrics.
The technological innovation benefits companies involved in industrial processes. It aims to improve working conditions, safety, and profits in this industry segment. To achieve these goals, monitoring, analysis, and prediction of the state of industrial processes must be carried out. This has become feasible with the rise of Industry 4.0, where machines and equipment can be continuously monitored using various sensors. Optimizing industrial processes has become one of the main pillars of an efficient and safe industry. Therefore, when sensor data indicates that a machine or equipment is heading toward failure, preventive maintenance must be carried out. This preventive step reduces downtime, prevents complete machine or equipment failure, and enables optimization of the entire industrial process. Unfortunately, sensor data alone is often insufficient to detect failure patterns, even with experts in the field. Thus, the use of machine learning methods on such sensor data is desirable and necessary to ensure unbiased recognition of patterns in the data. By using this method for predictive maintenance, the companies can optimize industrial processes and indirectly increase productivity and efficiency. This means that the method can be used for different cases (anomaly detection, reconstruction of distorted data, data reduction, and many others). The way it is used depends on user parameters and data. - Advantages and Innovations
-
• Innovative method for predictive maintenance in industrial processes uses advanced techniques inspired by nature and machine learning.
• The method automatically finds optimal anomaly detection models for each case, saving time and money.
• It allows companies to monitor and optimize industrial processes and increase productivity and safety.
• It is a flexible method that can be used for various sensors depending on user parameters and data.
• It automates the process of defining and constructing models, reducing the need for expert knowledge and subjective decision-making.
• It offers an objective solution to complex problems through computer algorithms and evaluation matrices. - Stage of Development
- Lab tested
- Sustainable Development Goals
- Goal 9: Industry, Innovation and Infrastructure
- IPR status
- Secret know-how
Partner Sought
- Expected Role of a Partner
- They are searching for companies who are interested in the chance to collaborate with the inventors and the university to obtain a license for the technology and further develop it. These partners could include businesses that manufacture sensors or those involved in predictive maintenance.
- Type and Size of Partner
- SME 50 - 249
- SME 11-49
- Big company
- Type of partnership
- Commercial agreement with technical assistance
- Research and development cooperation agreement
- Investment agreement
Dissemination
- Technology keywords
- 01003003 - Artificial Intelligence (AI)
- 01003012 - Imaging, Image Processing, Pattern Recognition
- 01003008 - Data Processing / Data Interchange, Middleware
- Market keywords
- 02007020 - Artificial intelligence programming aids
- Targeted countries
- All countries