{"id":35861,"date":"2025-07-18T10:27:33","date_gmt":"2025-07-18T08:27:33","guid":{"rendered":"https:\/\/risc.web-email.at\/referenzprojekte\/proof-of-concept-for-monitoring-the-flame-cutting-process\/"},"modified":"2026-03-10T14:22:40","modified_gmt":"2026-03-10T13:22:40","slug":"proof-of-concept-for-monitoring-the-flame-cutting-process","status":"publish","type":"project","link":"https:\/\/risc.web-email.at\/en\/referenceprojects\/proof-of-concept-for-monitoring-the-flame-cutting-process\/","title":{"rendered":"Proof of concept for monitoring the flame cutting process"},"content":{"rendered":"
\n
\n
\n
\n

In an innovative proof-of-concept project, RISC Software GmbH has developed a solution for the automated detection of nozzle wear in the flame cutting process for framag Industrieanlagenbau GmbH.<\/em><\/p>\n<\/div>\n\n\n\n

\n
\"\"<\/figure>\n<\/div>\n<\/div>\n\n\n\n

The flame cutting process is a thermal separation process that is used to cut metals. Nozzles are used to precisely guide the oxygen jet and mix the fuel gas evenly. framag uses the flame cutting process to cut materials efficiently and precisely in the steel industry. <\/p>\n\n\n\n

The aim of the project was to find out whether the condition of the nozzles – in particular the wear caused by deposits – can be recorded and analyzed using body emission data. This data, which was recorded during the cutting process, should provide information on whether it is possible to differentiate between new and worn nozzles. <\/p>\n<\/div>\n<\/div>

\n
\n

Process sensor data and uncover patterns<\/strong><\/h3>\n\n\n\n

To ensure efficient and accurate analysis, a key aspect of the project was to convert the raw data from the body emission sensor into an industry standard format. This sensor data acts as an indicator to better understand the differences in the behavior of the nozzles – a key indicator of their condition. This data was collected during a production-like process. After data processing, spectral analysis was used to uncover patterns indicative of nozzle wear. <\/p>\n\n\n\n

The results are promising. It was shown that the body emission data actually provides valuable information about the condition of the nozzles. Primarily, it was possible to recognize that wear is reflected in the data, which lays the foundation for a possible automation of wear detection. However, the project also brought some challenges to light. For example, the limited variety of nozzles proved to be a challenge, which will require an expanded database in future projects. <\/p>\n<\/div>\n<\/div>

\n
\n

Recognizing potential with machine learning methods<\/strong><\/h3>\n\n\n\n

Nevertheless, the proof-of-concept was an important step towards optimized, data-based monitoring of the flame cutting process. The knowledge gained from the RISC software will form the basis for further initiatives by further developing the models and verifying them with more data. In the long term, this solution could significantly increase the efficiency and precision of flame cutting by allowing maintenance to be carried out in a targeted and proactive manner. <\/p>\n\n\n\n

This project has shown that innovative technologies such as body emission sensors and machine learning methods have the potential to make industrial manufacturing processes more intelligent and efficient.<\/p>\n<\/div>\n<\/div>\n

\n
\n

Project partners<\/h2>\n\n\n\n
\"\"\/<\/figure>\n<\/div>\n\n\n\n
\n

Project details<\/h2>\n\n\n\n
    \n
  • Short title: <\/strong>Wear detection in the flame cutting process using machine learning<\/li>\n\n\n\n
  • Long title: <\/strong>Recognition of wear detection of nozzles of the flame cutting process by signal processing<\/li>\n\n\n\n
  • Project partner:<\/strong> framag <\/em>Industrieanlagenbau GmbH<\/li>\n\n\n\n
  • Duration:<\/strong> 08\/2024-10\/2024<\/li>\n<\/ul>\n<\/div>\n<\/div>\n\n
    \n
    \n

    Ansprechperson<\/h2>\n\n\n\n
    \n
    \n

    <\/p>

      <\/ul><\/div>\n
      \n
      \n<\/fieldset>\n
      \n\t
      \n\t\t

      \n\t\t

      \n\t
      \n\t\t

      \n\t\t

      \n\t
      \n\t\t