Saving Energy and Material Costs on the Basis of Big Data Analyses
By susanne.gutermuth@cosmino.de InSaving Energy and Material Costs on the Basis of Big Data Analyses
2022 is not an optimal year for the manufacturing industry, and 2023 threatens to be no better. Material and energy bottlenecks and cost explosions are the main causes. Reductions in consumption now seem necessary, no matter how great the dependence on gas or disrupted supply chains actually is. Above all, you have to look at these savings from an economic point of view: It has never been more worthwhile – while maintaining the same output – to reduce material, runtime and energy. Or to streamline the portfolio and put existing material and available energy into the most promising products. And perhaps there are free capacities of people and machines right now to carry out such measures.
Added Values of Big Data
By analyzing mass data (Big Data) you can, for example:
- identify the existing correlations between process parameters and output quantities/qualities, for example, to detect scrap as early as possible and save valuable machine runtime and material,
- determine the maintenance requirements of systems on the basis of the condition of wear parts in order to carry out preventive maintenance instead of repairing only in the event of damage,
- recognize unusually high energy consumption and find out its causes,
- identify cost-intensive processes and products and look for alternatives.
How Recorded Data is Used for Analysis
With Cosmino MES and Digital Twins, machine data can be collected in large numbers and – linked to order, product, machine status, quality – are available for analysis purposes, as we presented in this report . Big Data does not even enter the relational database of the MES, but is stored in specialized databases. Therefore, Digital Twins transfer the process data – after it has been visualized and checked against limits by the MES – into a time series database for further storage and analysis. To save IT costs, this can also be located in a cloud platform. Regardless of whether on-premise or in the cloud, in both cases sufficient analysis options are available to generate real added value and savings from this valuable treasure trove of data.
Most of the Data Required for Such Analyses Often Already Exist
A basic set of machine states, orders, quantities, quality, etc. is often already recorded by an MES or PDA. Process characteristics and energy values, if applicable, are often not yet in the same data pot or are missing entirely.
Such information that is now required, e.g. from machines or smart meters, can now be additionally collected with Digital Twins and integrated into the existing SFDC/MES environment or a database can be set up for analysis purposes. On the Digital Twin, the process and energy data are supplemented with information from the store floor data collection or the MES. This could be – depending on availability and demand – the recorded tool, product, material, order, machine status, quality status and batch number.
We Get Added Value for You from Your "Big Data" – for Example, Through Machine Learning
Big Data specialists can then analyze the use cases defined by process experts. The first step is to check whether certain constellations have an influence on the use case. For example, whether the curves of energy consumption or process parameters show wear of parts or the manufactured article. If an influence is detected, the algorithm found can be "trained" to identify the defined cases in the future with high probability from the recorded live data.
It is also possible to define maximum limits for the energy consumption of individual machines, components or areas, and if the consumption is too high, an algorithm informs the user – based on which areas are actually producing and thus consuming energy.
Tremendous Opportunity to Find Significant Savings with Big Data
The fact is that no process today already works perfectly. Energy, material and runtime is wasted everywhere, consciously or unconsciously. Even a few found cases can generate large sums of savings, which is why an investment in such an acquisition and analysis project almost always turns out to be extremely worthwhile.
According to a research project, savings of almost 1 billion € and an average saving of 2.87% of the sales volume of the affected process were determined for a considered number of 338 Big Data projects from 1994 to 2020. The costs for software and service were far below this, so that on average a payback was achieved after just 15 days.
We – COSMINO AG – would be pleased to support you. Be it with an acquisition concept, software or the analysis and development of Big Data scenarios. Better today than tomorrow, decide on an initial consultation to explore the possibilities. We look forward to hearing from you.