Falling margins and rising energy prices are a particular burden for those companies in the manufacturing industry who cannot make their products without high energy consumption. Octotronic has developed a digital twin that enables companies to model their energy requirements and show them for each individual product. The software utilises data from the production process, along with machine-learning procedures.
Picture: Octotronic
One might think it would be easy to calculate a company’s energy consumption. In reality though, this is a rather complex endeavour and energy consumption can fluctuate significantly. It can be affected by factors such as room and/or ambient temperatures, failure to observe warm-up, balance or idle times, and incorrect or inadequate insulation in machinery.
Climate neutrality, rising energy costs and falling margins can make it worthwhile for companies in the manufacturing industry to take a closer look at their energy flows: be it to identify potential savings, to calculate actual production costs or to draw up an energy requirement plan. Given the need to decarbonise entire sectors of the economy, and in view of rising energy costs, accurate analysis of energy flows is becoming increasingly important.
To determine electricity consumption more precisely, Octotronic develops a digital twin that is tailored to each customer’s requirements, in order to determine the energy flows. This so-called ‘energy twin’ is able to model the energy consumption of a company or factory and show it broken down into three categories: product, processes and infrastructure. The energy twin utilises existing data on one hand, for instance from ERP (enterprise resource planning) systems, and, on the other hand, measuring devices that measure electricity consumption or ambient temperature. At product level, each individual manufactured product is given a virtual electricity meter. This meter measures how much energy is consumed for production at each stage of the process. This measurement data is isolated and adjusted: Influencing factors, such as process parameters and ambient temperature, are mathematically filtered out, so that the data can be considered separately.
At process level, the amount of energy required to start machinery is analysed, for instance based on warm-up times. At infrastructure level, this data can then be compared with other energy flows for lighting, heating and/or cooling.
Algorithms and AI components can be used to analyse and model consumption figures. This means, for example, that it is not necessary to measure the energy consumption of all machines of a certain type at all times, so the number of measuring devices can be significantly reduced. Moreover, it is also possible to calculate energy consumption even if production takes place at different locations.
Substantial cost savings can be made with the aid of such a solution, especially by companies operating in sectors where margins are low and energy costs are high. In addition, an energy twin makes it possible to determine the carbon footprint of an individual product, which can drive a transformation towards more sustainable production methods. After all, consumption has to be monitored on an ongoing basis if constant improvements are to be made.
In practice, however, there are numerous evident challenges that have to be overcome before factories can be interconnected. Data that is stored in programmes with no option for connecting to external data, i.e. in data silos, must be made accessible and processed accordingly.
Such digitalisation projects for interconnecting production machinery (see the article on connected machines) also require a transformation of company processes.
From a technical perspective, it is often clear what needs to be done. Nevertheless, the integration of digital twins requires a clear strategy that does not start with solutions, but instead takes problems that are waiting to be solved as its starting point.
