Mit dem Laden des Videos akzeptieren Sie die Datenschutzerklärung von YouTube.
Mehr erfahren
Introduction Data Science
Are you already using the full potential of your data? Which technical or business added values can be derived from it? What patterns and anomalies do your data contain? Which correlations and predictions can be derived from your data?
We at IQZ can help you answer these and other questions by analyzing your data. The catchwords industry 4.0, big data, short development cycles and agility in development are very likely to be mentioned today when challenges are addressed by companies. These can only be overcome by using data science if all available data is used systematically. With our expert knowledge we analyse your data and help you to use this data profitably for your company.
Our offer:
- Initial workshops and check of your data structure
- Determination and visualization of key figures
- Data-supported cause analysis
- Feedback of field information into the development process
- Information generation for new business models (e.g. Pay per Use, Predictive Maintenance etc.)
- Recognition of anomalies and patterns (e.g. in warranty management)
- Analysis of real-time data from the field
- Modeling of reliability scenarios
- Creation of forecasting models
- Production process optimization based on process and machine data
The following methods are used in extracts by us:
- Regression analyses and parameter reduction methods (e.g. Shrinkage, Ridge, Lasso)
- Analysis of Variance (ANOVA)
- Visualization procedures
- Neural Networks
- Machine learning
- Bayesian statistics
- Simulation
- Stochastic Models
- Optimization procedures
What is data science?
The quality of products is a characteristic which customers consciously perceive and which can make a decisive decision about the purchase of a product. If quality is poor, poorly quantifiable risks such as loss of image or negative reporting are particularly critical. Consequently, it is extremely important for companies to ensure a high level of quality even with increasing product complexity and increasing price pressure. This can be achieved with the help of data analytics. Data analytics refers to the scientific process of extracting and analyzing data. The aim is to determine conclusions or correlations by evaluating existing data. For example, the analysis of existing warranty data can help to identify future quality problems in the field ahead of time or to identify commonly occurring damages.
The following figure shows the impact of different production lines of the same product on the service life. This can be used, for example, to discuss process optimizations or to determine clusters for field actions.
The field of data analytics is broadly diversified. The creation of reports, key figures, drilldowns and alerts provides access to and an overview of a large data volume. This is of particular interest, for example, to determine the time, background and frequency of certain events from the existing data. In addition, statistical methods can be used to identify the underlying causes of deviations or anomalies in the data volume. Forecasts and optimizations can be used to make predictions with a confidence level or to determine trends. Answering these questions with the help of data analytics offers new opportunities to control, monitor and optimize data-driven processes. As the complexity of the questions and the amount of data increase, so does the complexity of the analysis. We use a variety of methods to meet this challenge with you.
The next figure shows a scatterplot matrix which shows the correlation between various influencing factors on a measurable target value. This allows the visual recognition of patterns in the influence factors. In addition to the graphical evaluation, a mathematical weighting of the influencing factors also takes place.
Portfolio of methods
Neural networks
Artificial neural networks (ANN) are used to design and solve models. They are able to generalize abstract structures and identify patterns, even with incomplete or fuzzy data. The actual model should be seen as a black-box within which the model is mapped from the input parameters to the solution via artificial neurons and their links, thus explaining the term neural networks.( refer to following figures: Neural networks)
Figures: Neural networks
ANN are particularly useful if the problem is very complex and the mathematical description (model) as well as the solution to the problem are practically impossible (no knowledge of the influencing variables and the functional relationships). Typical application cases for ANN include parameter estimation (e.g. the adjustment of a multi-parameter Weibull function), lifetime prediction (e.g. taking into account climatic influencing variables) through to the prediction of share prices.
Selected references:
- Lifetime prediction taking into account changing climatic conditions.
- Neuronal parameter estimation for the use of failure data
- Neuro-fuzzy reliability optimisation method with respect to life cycle costs
- Optimisation and vague date in the development process
Your challenge:
- Were you unable to obtain a satisfactory result with conventional methods, no data and bases to design models were available?
We offer you the mathematical description and ANN-based solutions to your individual questions – get in touch, we will be glad to help.
Fuzzy logic
Fuzzy logic enables the subjective knowledge that is available in the organisation or from experts to be used and integrated in existing approaches to safety and reliability analyses (refer figure:Fuzzy logic). A further advantage is that the fuzziness and subjectivity of the input data can still be identified in the results of the analysis.
Figure: Fuzzy logic
Fuzzy logic has proven its worth in safety and reliability analyses in combination with tools such as FMECA or FTA to integrate fuzzy data (e.g. failure rates). The fuzzy data is initially fuzzified (transposed to a set framework of fuzzy mathematics). The result is then generated according to the calculation rules of the superordinate tool (FMECA, fault tree, etc.), the following defuzzification of the result serves to transpose the fuzzy result into a concrete value.
Selected references:
- Fuzzy FMEA to analyse the risk from a pressure tank system
- Fuzzy fault tree analysis for a vehicle sensor system
- Methodical comparison of the Boolean and fuzzy error tree analysis taking a flight control system as an example
Your challenges:
- Does your system contain uncertainties, are you unable to identify concrete input variables (e.g. probabilities of occurrence, probabilities of detection, failure rates)?
- Are the experts who have been consulted unable to agree on the identification if the input variables (e.g. probabilities of occurrence, probabilities of detection, failure rates)?
We can offer the necessary tools, help you with their application and help you prepare the necessary documents – get in touch, we will be glad to help.
Get in touch with us!
You’ve got questions about Data Science?
Feel free to contact Andreas Braasch!
Telefon: +49 (0)202 – 515 616 90
Mail: info@iqz-wuppertal.de
