AI-powered laser welding from TRUMPF boosts auto manufacturing quality
Technology

AI-powered laser welding from TRUMPF boosts auto manufacturing quality

TRUMPF is using artificial intelligence (AI) to increase productivity in automotive manufacturing. The high-tech company has developed an AI solution that checks the quality of components, such as batteries for electric cars, immediately after laser welding, identifies possible defects and enables repairs in the laser welding station. "This enables users to reduce their manufacturing costs because the AI solution can replace other time-consuming quality inspections," says Martin Stambke, the TRUMPF product manager responsible for AI quality inspection. The first customers from the automotive industry are already using the solution in series production and the high-tech company is now presenting it at its in-house exhibition INTECH AI solution suitable for many applications

The solution takes a picture of the weld seam and AI analyses whether it meets the specifications defined by the user for the manufacturing process. The sensor technology generates traceable measurements for the user, enabling transparent quality control without black-box decisions. For example, if the weld is too narrow or too wide, the AI notifies the user while the component is still in the laser cell.

The operator can quickly decide if a repair is necessary, thus ensuring the quality of the weld. This increases first-pass yield, the proportion of components that meet quality standards on the first pass. "This is particularly important for manufacturers of batteries for electric cars, for example, who will have fewer rejects and thus reduce costs," says Stambke. The AI solution is suitable for many other applications, such as welding body parts, hairpins for electric motors or contacts for electronic components. No programming knowledge required

“Customers get an easy-to-use AI tool that integrates seamlessly with the TRUMPF system of lasers, sensors and optics," says Stambke. Users can use the TRUMPF AI without any programming knowledge. All you have to do is train the artificial intelligence before using it. To do this, the user marks correct and incorrect welds in the corresponding image material. For the training process, TRUMPF offers EasyModel AI, a software program that is as easy and intuitive to use as a smartphone app. Only a few training images are required.

TRUMPF is using artificial intelligence (AI) to increase productivity in automotive manufacturing. The high-tech company has developed an AI solution that checks the quality of components, such as batteries for electric cars, immediately after laser welding, identifies possible defects and enables repairs in the laser welding station. This enables users to reduce their manufacturing costs because the AI solution can replace other time-consuming quality inspections, says Martin Stambke, the TRUMPF product manager responsible for AI quality inspection. The first customers from the automotive industry are already using the solution in series production and the high-tech company is now presenting it at its in-house exhibition INTECH AI solution suitable for many applications The solution takes a picture of the weld seam and AI analyses whether it meets the specifications defined by the user for the manufacturing process. The sensor technology generates traceable measurements for the user, enabling transparent quality control without black-box decisions. For example, if the weld is too narrow or too wide, the AI notifies the user while the component is still in the laser cell. The operator can quickly decide if a repair is necessary, thus ensuring the quality of the weld. This increases first-pass yield, the proportion of components that meet quality standards on the first pass. This is particularly important for manufacturers of batteries for electric cars, for example, who will have fewer rejects and thus reduce costs, says Stambke. The AI solution is suitable for many other applications, such as welding body parts, hairpins for electric motors or contacts for electronic components. No programming knowledge required “Customers get an easy-to-use AI tool that integrates seamlessly with the TRUMPF system of lasers, sensors and optics, says Stambke. Users can use the TRUMPF AI without any programming knowledge. All you have to do is train the artificial intelligence before using it. To do this, the user marks correct and incorrect welds in the corresponding image material. For the training process, TRUMPF offers EasyModel AI, a software program that is as easy and intuitive to use as a smartphone app. Only a few training images are required.

Related Stories

Gold Stories

Next Story
Resources

Schindler named among Forbes’ World’s Best Employers 2026

Schindler has been named among Forbes’ World’s Best Employers 2026, a global ranking recognising companies for their workplace environments.Compiled by Forbes in partnership with Statista, the ranking is based on independent employee feedback, including whether employees would recommend their organisation to friends and family.The company said the recognition reflected its focus on creating a workplace where employees can develop their expertise and build careers. ..

Next Story
Infrastructure Urban

India needs an export-execution mindset: Vishesh Jain

Interview with Vishesh Jain, Founder & CEO, EximpleWhat gaps in India’s global trade ecosystem led you to build Eximple, and what problem did you see businesses struggling with most?Eximple came out of a problem that all three founders had seen from different choke points in the trade ecosystem. India has millions of MSMEs and manufacturers, and many of them can make products that are genuinely competitive on quality. Yet only a small fraction is able to build meaningful international businesses. In our experience, the gap is usually not manufacturing capability; it is the ability to exe..

Next Story
Building Material

Nuvoco reports 13.5% revenue growth in Q2 FY27

Nuvoco Vistas Corp. Ltd. reported consolidated revenue of Rs 27.9 billion for Q2 FY27, up 13.5% year on year. Cement sales volume rose 9% to 4.7 million tonnes (MMT), while EBITDA increased 11% to Rs 4.11 billion.The company said it began its cement operations in 2014 with a 2.3 MMT per annum (MMTPA) operation at Nimbol. It now aims to reach 35 MMTPA of cement capacity by FY28, nearly 15 times its initial capacity. Its portfolio spans cement, ready-mix concrete and modern building materials.On sustainability and cost efficiency, Nuvoco partnered with CleanMax for a 46.4 MW wind-solar hybrid re..

Advertisement

Subscribe to Our Newsletter

Get daily newsletters around different themes from Construction world.

STAY CONNECTED

Advertisement

SPECIAL OFFER
QR Code