In mining, tool wear is a daily occurrence – the teeth of excavator buckets, chisels, crushers and other components are subjected to extreme mechanical and thermal stresses. Until now, the approach to their maintenance has been based on replacing worn parts and remelting them into new ones, with high material, logistical and environmental costs. Today, thanks to a combination of laser material deposition (LMD) technology and artificial intelligence, this scenario could change dramatically.
The AI-SLAM (Artificial Intelligence Enhancement of Process Sensing for Adaptive Laser Additive Manufacturing) project, carried out by the Fraunhofer Institute for Laser Technology ILT in collaboration with Canadian partners, represents a milestone in the automation of metal component repair. The technology, until recently the domain of highly specialised engineering teams, can now operate almost autonomously – at the touch of a button.
Laser material deposition – precise layer-by-layer remanufacturing
LMD is a surfacing process that uses a focused laser beam to deposit metal powder (such as stainless steel and tungsten carbide) onto worn components. This process not only restores the original shape of the tool, but also makes it possible to locally strengthen the most stressed areas without having to replace the entire part.
During surfacing, the laser melts the stainless steel at ~1300°C, creating a weld pool. At the same time, the nozzles emit tungsten carbide particles (melting point ~2900°C), which penetrate the molten material and solidify to form a hard, wear-resistant coating. It is important to maintain the correct ratio of these two components – an excess of carbide causes brittleness, a deficiency reduces durability. Until now, precise control of process parameters, from laser power to tracking speed, has required manual monitoring. This is now taken over by AI.
Artificial intelligence as process operator
A new artificial intelligence module developed at the ILT can automatically analyse, plan and monitor the tool reconditioning process. The module consists of several functional layers:
- Tool geometry mapping: A CMOS camera scans the worn component and creates an image of its current shape.
- Comparison to a reference model: The software analyses the differences between the scan and the reference model of the new part.
- Calculate paths and surface parameters: AI determines where and how much material to deposit, taking into account layer thickness, temperature, alloy composition and surface slope, among other factors.
- Real-time control: During the process, images from the camera are fed back to the AI system, which detects and corrects any deviations from the programmed trajectory.
This is a breakthrough as the management of the plating process involves up to 150 key parameters, which are now set dynamically without human intervention.
The AI-controlled LMD process simplifies the role of the operator when coating excavator teeth / © Fraunhofer ILT
Technology ecosystem: an international partnership
The AI-SLAM project is an example of a fruitful collaboration between research institutions, universities and industry:
- The Fraunhofer ILT is responsible for the development of the LMD technology and the integration of AI into the surfacing system.
- The National Research Council of Canada and McGill University are leading the R&D work.
- Braintoy (Calgary) is providing machine learning algorithms and an analytics platform.
- Apollo Machine and Welding (Edmonton) is implementing the technology in an industrial setting.
- BCT (Dortmund, Germany) has integrated an AI module with OpenARMS software to enable automatic machine control without the need for G-code programming.
With this integration, shop floor operators no longer need to manually configure machines. All they have to do is load a component, press ‘start’ – and the system carries out the repair itself.
Faster, more efficient, greener
The benefits of implementing an AI-SLAM solution are manifold:
- reduction of repair times by up to 50%,
- elimination of human error,
- savings in raw materials and energy – instead of manufacturing new components,
- we regenerate existing ones, lower operating costs and less industrial waste.
What’s more, the technology is not limited to the mining industry – its potential applications include the energy, engineering and agricultural industries, as well as aviation and heavy transport, among others.
