Industrial Robots Welding Automation and Optimized Programming Times

Related Vendors

FAULHABER Drive Systems

Wuxi InfiMotion Technology Co., Ltd.

The introduction of Visual Components OLP offline robot programming has fundamentally changed the entire welding process at Sandvik Turku. While programs are created offline, the robot cells remain productive. The result: no lost production time and significantly higher capacity utilization.

Sandvik Mining produces robust underground loaders and dump trucks in Turku, Finland, that are used under extreme conditions. The machines are made of massive main frames, lift arms, and buckets that need to move heavy loads—often in tight tunnel environments. Accordingly, the welds and assemblies used in manufacturing are quite complex.

To program and efficiently automate these demanding structures, Sandvik has been relying on industrial robots for years. However, manual programming of the complex, multi-layered welds has proven to be a bottleneck: This approach takes a long time, requires extensive trial and error, and leads to unnecessary downtime in the robotic cells. Additionally, it is difficult to predict whether certain welds are even accessible—complicating the planning process and limiting the utilization of the welding robots. With the introduction of the Robot Offline Programming (OLP) software from Visual Components, Sandvik is accelerating programming, improving weld quality, and significantly advancing automation.

Multi-Layer Welding in Massive Assemblies: High Complexity, High Effort

The plant in Turku is part of Sandvik Mining's Load and Haul Division and has been manufacturing underground machines for global applications in mining and tunneling since the early 1980s. Around 900 employees produce all main frame components, including front and rear frames.

The requirements for these components are high: a single front frame consumes over 100 kilograms of welding wire (~220 pounds), and most of the welds are multi-layered. This necessitates hundreds of weld beads that must be precisely programmed and executed in the optimal sequence. Additionally, Sandvik prefers to weld in the PA position, which adds extra effort for irregular bevels.

Because employees previously had to manually test every possible robot position without knowing whether the weld was accessible, downtime occurred, which directly impacted productivity and throughput. At the same time, it became increasingly difficult to plan optimal sequences or heat inputs to avoid distortions.

Virtual Programming for Higher Efficiency

To address these challenges, Sandvik is utilizing offline programming with Visual Components. The software allows for virtual planning, simulating, and validating welding paths—before a robot is even positioned in the cell. By simulating the welding processes, employees can better organize manufacturing sequences, optimize critical welding positions, and reduce undesirable distortions. The effect is particularly evident with larger assemblies, such as the front frames: What previously required days or weeks of teach pendant programming can now be prepared in just a few days—with fine-tuning in the cell often taking only a single shift.

Another advantage is standardization: Visual Components OLP enables multi-layer welding profiles that can be applied to different shapes—such as curves or plate edges. In addition to pure path planning, Sandvik also uses Visual Components for the preparation of fixtures and clamping devices. The team creates point clouds of the burner paths and analyzes the optimal orientation of the frame assembly concerning the welding positioner. The resulting boundary conditions are directly incorporated into the design of the fixtures.As a result, Sandvik was able to develop a fixture that ensures full accessibility to all weld seams while simultaneously enabling ideal welding positions. This significantly reduces the effort in the early project phases—and also minimizes the risk of later adjustments.

Faster Production Start-Up and Higher Robot Availability

With the introduction of Visual Components OLP, the entire welding process at Sandvik Turku has fundamentally changed. While programs are created offline, the robotic cells remain productive. The outcome: no lost production time and significantly higher utilization.
 For particularly large frame components, the production start-up is now completed within a few weeks. Smaller projects are ready for programming within a few days, and the final adjustments often take only one shift. The welding quality has also greatly benefited from the standardization of multi-layer profiles. The processes are more reproducible and significantly more uniform across various assemblies.

With offline programming, we can check the accessibility of each weld seam in advance, thereby avoiding wasting time without results.

Strengthening Design for Manufacturing

For the future, Sandvik plans to further integrate the capabilities of offline programming into the design process. By validating accessibility and weldability early on, designers can develop more efficient designs and make decisions that simplify automation even further. Based on this, Sandvik aims to continue increasing productivity at the Turku plant and to consistently pursue the path towards a harmonized, highly automated welding manufacturing process.
 "The programming of an entire front frame using the teach pendant takes a long time, and valuable production time is lost during this process," explains Markus Juntunen, Production Development Engineer at Sandvik. "With offline programming, we can check the accessibility of each weld seam in advance, thereby avoiding wasting time without results. Additionally, it's much easier to standardize the parameters of the burner position: We create multi-layer profiles that we can quickly transfer to different bevels or geometries. This allows us to achieve reproducible quality in a short time. And when we set up a new robotic cell, we can significantly expedite the ramp-up process thanks to Visual Components."

*Business Development Director Robotics at Visual Components