| Title: | Point cloud-based robot welding trajectory generation for half-open corner joints |
|---|
| Authors: | ID Karner, Timi (Author)
ID Belšak, Rok (Author)
ID Gotlih, Janez (Author)
ID Klančnik, Simon (Author) |
|---|
| Files: |  https://www.sciencedirect.com/science/article/pii/S0921889026002848?via%3Dihub
 1-s2.0-S0921889026002848-main.pdf (13,17 MB)
MD5: D0C14F6D970A562D039C8D470533D29C
|
|---|
| Language: | English |
|---|
| Work type: | Article |
|---|
| Typology: | 1.01 - Original Scientific Article |
|---|
| Organization: | FS - Faculty of Mechanical Engineering
|
|---|
| Abstract: | Robotic welding for complex geometries, such as the half-open corner joint, requires precise, automated trajectory planning, a challenge largely unaddressed in current literature. This paper presents a novel point cloud-based method for automated robot welding trajectory generation specifically for this joint type. The innovation is a complete and computationally efficient pipeline, leveraging 2D laser line sensor data and point cloud processing to automatically detect the joint and generate an optimized, geometry-consistent 6-DoF trajectory. Crucially, the method calculates the optimal welding torch orientation, consistently maintaining an inclination relative to the joint geometry to ensure weld quality. The generated trajectory is automatically converted into executable ABB RAPID code. In simulation, the approach demonstrates high precision against a CAD reference, achieving a mean trajectory error of only 0.450 mm and a maximum error of 1.033 mm, enhancing automation and efficiency in challenging robotic welding applications. The tool center speed comparison between generated trajectory with algorithm and programmed trajectory also shows a difference of less than 0.17 mm/s. It has also been shown that the generated trajectory with algorithm consumes less energy than the programmed trajectory. |
|---|
| Keywords: | robotic welding, 2D laser sensor, point cloud, half-open corner joint, trajectory generation |
|---|
| Publication status: | Published |
|---|
| Publication version: | Version of Record |
|---|
| Submitted for review: | 22.12.2025 |
|---|
| Article acceptance date: | 29.06.2026 |
|---|
| Publication date: | 03.07.2026 |
|---|
| Publisher: | Elsevier |
|---|
| Year of publishing: | 2026 |
|---|
| Number of pages: | 14 str. |
|---|
| Numbering: | Vol. 205, [article no.] 105612 |
|---|
| PID: | 20.500.12556/DKUM-98753  |
|---|
| UDC: | 621.71:007.52 |
|---|
| ISSN on article: | 1872-793X |
|---|
| COBISS.SI-ID: | 283708419 |
|---|
| DOI: | 10.1016/j.robot.2026.105612 |
|---|
| Publication date in DKUM: | 06.07.2026 |
|---|
| Views: | 448 |
|---|
| Downloads: | 18 |
|---|
| Metadata: |    |
|---|
| Categories: | Misc.
|
|---|
|
:
|
Copy citation |
|---|
| | | | Average score: | (0 votes) |
|---|
| Your score: | Voting is allowed only for logged in users. |
|---|
| Share: |  |
|---|
Hover the mouse pointer over a document title to show the abstract or click
on the title to get all document metadata. |
