AirTwin DPP: AI-powered digital twin for pneumatic systems

AirTwin DPP develops an AI-powered digital twin for pneumatic systems in machine tools. It detects and localizes compressed-air leaks in real time and reduces consumption by 5–10%. Via the Asset Administration Shell and Digital Product Passport, it generates standardized, auditable energy and CO₂ data.

Key data

Description

Background

Compressed air is one of the most important energy sources in industrial manufacturing. In machine tools, it powers clamping, cleaning, and tool-change functions. At the same time, it is highly inefficient: leaks often go undetected and cause significant energy losses and indirect CO₂ emissions. Today, such leaks are usually located by hand. This costs time and expertise.

Objectives

The project develops a digital twin for pneumatic systems. A digital twin is a virtual model of the real machine. Artificial intelligence (AI) detects leaks in real time and shows where they occur — down to individual components. The goal is to reduce compressed-air consumption by 5 to 10 percent. The project also generates reliable data on energy use and CO₂ emissions.

Approach

Sensors capture pressure and flow signals directly on the machine. AI models distinguish real leaks from normal fluctuations during operation. An AI-assisted commissioning function automatically checks control parameters during setup and after every component replacement. All data is stored in a standardized way. For this, the project uses the Asset Administration Shell (AAS) and the Digital Product Passport (DPP). These standards make the data interoperable and auditable. The solution is tested on two machine tools under real production conditions.