New potentials for improving efficiency in foundries
The efficiency boost achieved using Tvarit's new TiA platform rests on four fundamental principles. The focus here is on reducing rejects and reworking, facilitate energy management on part level to cut energy losses, enhancing system availability and improving maintenance.
Expert knowledge of processes and products can be secured and expanded by employing innovative software technologies such as Finite Element Analysis and AI.
In parallel to further automation of production processes in foundries using TiA, Tvarit's aim is also to provide the user with an efficient, easy to operate tool for automated evaluation of large volumes of data in real time and for predictive process control.
The German foundry industry in flux
The German foundry industry is currently confronted with numerous challenges in assuring the future of the relevant companies.
The challenges of climate change, involving rising costs for assuring sustainability and decarbonisation, steeply rising energy, and feed materials costs, as well as the effects of the current crises, hit the foundry industry, with its only modest profit margins on its products, particularly severely.
The situation is then further exacerbated by the current trend in the automotive industry toward the abandonment of the internal combustion engine and the adoption of alternative propulsion systems, combined with the need to master an ever-increasing range of variants.
Many companies appear to have reached their operational optimum of energy-intensive production in the high-wage location of Germany by means of modern management methods such as continuous improvement programs, lean management, and similar innovations.
A further series of problems arises from the fact that more and more expert employees are now retiring and that their companies then encounter and have to deal with significant skill and qualification deficits, a task which is proving increasingly difficult.
The product-development process has been effectively digitalised in many foundries. It is then more astounding that the great economic and quality potentials that can be realised by means of digitalisation of production, in particular, have up to now scarcely gained significance.
Many foundries are deterred by incalculable investment expense and the risks encountered on the road to digitalisation. In many cases, attempts are made to achieve improvements in the process route via isolated investments in individual machines and software tools. The result in recent years has frequently been a highly heterogeneous IT landscape. Many companies have differing systems in use for differing applications. The software used and the solutions developed by in-company IT departments are in many cases not totally compatible with operational procedures and are over-dimensioned. This makes the systems too complex for the user and their use inefficient; operating potentials can thus be realised only with difficulty.
Large quantities of data are in many cases acquired using the system landscape, with its prompt evaluation involving an enormous amount of effort. At the same time, the inhomogeneity of the systems acts like a firewall between the various departments and the various working procedures. And there is a danger of loss of information and inefficiency when the staff in the various departments are unable to work smoothly together.
There are close links between the individual cost-influencing factors of productivity, quality, maintenance and energy consumption; these necessitate a uniform mode of analysis.
The methodology applied up to now fails to reveal many potentials and concealed losses. New conceptual solutions are needed to address this.
Successful expansion of digitalisation to achieve long-term improvements in efficiency necessitates extremely accurate and detailed analysis and description of the foundry processes and their evaluation, wherever possible, in real time. The user must in this context be provided with a tool that will enable him to quickly generate the necessary information, process-control proposals and forecasts.
The solution developed must permit the complete description of the influencing factors in a uniform system and it must also integrate smoothly into the existing organisational structure and working procedures.
The steadily rising demands made on castings in terms of process-certain mastery of high complexity in combination with high mechanical-strength requirements can be successfully achieved only by means of complete control and transparency of the processes involved. It is vital to keep the processes within narrow limits for production of thin-walled components for lightweight engineering, in particular.
Bionic designs that exhaust the potentials of dimensions and materials possess a significantly lower tolerance for deviations than conventional types.
A further objective of digitalisation can in many companies also be found in the perceptions contained in the intrinsic knowledge possessed by the employees, in combination with process data, which in fact constitutes the essential capital asset of the company - expert knowledge that needs to be treasured and increased. The target here must be the development of a future-viable mode of working that is less dependent on individual experience.
TiA - an AI-based technology for foundries
Our conceptual solution for detection of losses at our customers' facilities consists of the merging of data from various systems and of data from the machinery and other equipment combined with the analysis of this data in real time via the use of a hybrid AI technology.
The causes of existing deviations, on the one hand, and proposals for process improvements, on the other, are thus supplied to the user. In the foundry process, in particular, with its very numerous and in some cases non-quantifiable influencing factors, AI technology is able to develop to the full its strengths of multiple evaluation of large volumes of process data. The integrated learning effect results in the continuous improvement of prediction accuracy.
Since the priorities and technical preconditions in foundries differ very greatly, procedures and applications are individually matched to each customer.
Tvarit's experts carefully plan the deployment on the basis of understanding of the processes and existing organisational structures and working procedures in the particular foundries. The successful use of the technology is validated by means of a test phase conducted on a machine or system on which the initial potentials become apparent after only a few weeks, before the customer decides on series-installation.
In line with the recognisable success factor, Tvarit can point now up to no less than fifty successfully completed projects. One of those is with a world renowned & modern LPDC foundry customer. The subject of energy costs arose at one of its plant’s locations. During the production process, temperature and pressure sensors, a total of 80+ parameters, were polled at sampling rates, in some cases, up to one per second - for thousands of casting components manufactured. An engineer would have been hopelessly overburdened with this task, but only such detailed analyses make it possible to reveal previously undetected losses and potentials. An AI solution optimised for foundries enabled the company to reduce energy consumption by more than 20% per aluminium casting component. This results in annual energy-cost savings of several hundred thousand euros for the plant as a whole.
Many other highly efficient foundry companies, intend now to reduce rejects, energy consumption and emissions even further and put their faith for this in Tvarit technology.
The solution: Tvarit's AI platform
Like many parts of the processing industry, the foundry sector is a technologically demanding field that is highly dependent on innovation.
The large number of measurable and non-measurable influencing parameters found in foundries, necessitates a new and practically orientated methodology to achieve successful digitalisation.
Operational excellence and productivity are the secrets of success. But when a foundry is already at optimum operation, only artificial intelligence can detect the last potentials for further optimisation.
AI specialist Tvarit has put together a package of AI tools specifically for foundries. Using the TiA for Casting platform, savings potentials can be not only detected, but also realised. In TiA for Casting, the data experts supply the only holistic AI platform that addresses all operating losses in foundries.
This solution comprises a total of four independent modules:
Prescriptive Quality (PsQ) for reduction of rejects along the entire process chain.
Prescriptive Maintenance (PsM) for predictive maintenance, detection and elimination of process deviation.
Prescriptive Production Planning (PsP) for optimisation of production planning.
Prescriptive Energy Optimisation (PsE) for the reduction of energy losses from electricity, coal and gas consumption.