· Deployment & production· 日本語で読む
Comau and Fincantieri Field-Test MR4Weld Mobile Welding Robot
IFR reported on MR4Weld, an outdoor welding automation system developed by Comau and Fincantieri. The system mounts a 6-axis robot on a tracked mobile platform and identifies weld locations and sets paths. Its design features include adapting to variations in welding conditions and a method that lets workers guide the robot.
What happened
On November 12, 2026, IFR reported on MR4Weld, an outdoor welding automation system developed by Comau and shipbuilder Fincantieri. After field testing, the system was upgraded to use a higher-payload robot. IFR said it improved quality, performance and worker well-being in labor-intensive welding tasks.
Why it matters
In the setup described by IFR, workers specify the start and end points of a weld, while the robot identifies the joint and executes the path. A key question for welding in the field is how much the combination of a mobile platform and automatic condition setting can reduce reliance on specialized robot operation.
Analysis
A key feature of MR4Weld’s design is that it brings a robot mounted on a mobile platform to the work site rather than confining welding to a fixed installation. IFR describes a system that scans the joint before welding to determine factors such as the number of passes, then uses vision to follow the specified path. For work where the shape of the object and material conditions vary from one location to another, the ability to interpret conditions at the site, as well as mobility, is likely to be important operationally. However, the article excerpt does not provide figures for compatible materials or plate thicknesses, payload, welding speed or quality assessments. The claim of improvement is also reported by IFR, and the comparison conditions and evaluation method are not clear.
In related coverage, we reported on Comau’s efforts to incorporate NVIDIA computing and simulation technologies into robot controllers and design processes Comau integrates NVIDIA technology into industrial robot products and design processes. That article covered the link from simulation to factory deployment, while MR4Weld performs welding at the work site. There is no basis to consider the two part of the same development plan, but they can be read alongside each other as examples of different stages: robot design and validation, and application to work on site. MR4Weld also shares with EDP tests Comau’s mobile robotic factory for solar power plant construction the use of mobile automation in different work environments, although MR4Weld targets welding at shipyards.
From an industrial perspective, a point of interest is the division of labor in which workers specify the start and end points while the robot determines the welding path and process conditions. The meta-language interface also appears intended to enable operation without requiring specialized robot programming skills, but the article excerpt does not specify the range of instructions it supports or how it handles misrecognition. Details to watch for include the scale of field testing, quality and stoppage rates during continuous operation, worker safety, applicable welding conditions and the programming burden after deployment. Without this information, it is not possible to assess any productivity or cost advantage over fixed installations.
Sources
Translated from the Japanese original: /news/comau-to-fincantieri-yousetsu-you-idou-robotto-mr4weld-wo-ji-5d8cfd. The Japanese version is authoritative if they differ.