Offline robot programming assisted by task demonstration: an AutomationML interoperable solution for glass adhesive application and welding

TL;DR

This paper tackles the barrier of programming robots for non-experts in heterogeneous manufacturing environments. It proposes an intuitive offline programming workflow that captures operator skills from single-shot demonstrations using a magnetic tracker, fuses these with CAD/CAM positional data, and validates paths in simulation before generating programs. Data interoperability is achieved through PathML, an AutomationML-based syntax that unifies robot, process, and geometry data within a CPPS. Experiments in glass adhesive application and TIG welding show robot paths with positional errors up to $4$ mm and orientation errors between $1^ ext{$ ext{°}$}$ and $3^ ext{$ ext{°}$}$, within functional tolerance, demonstrating the method's practicality and potential for SME deployment.

Abstract

Robots have been successfully deployed in both traditional and novel manufacturing processes. However, they are still difficult to program by non-experts, which limits their accessibility to a wider range of potential users. Programming robots requires expertise in both robotics and the specific manufacturing process in which they are applied. Robot programs created offline often lack parameters that represent relevant manufacturing skills when executing a specific task. These skills encompass aspects like robot orientation and velocity. This paper introduces an intuitive robot programming system designed to capture manufacturing skills from task demonstrations performed by skilled workers. Demonstration data, including orientations and velocities of the working paths, are acquired using a magnetic tracking system fixed to the tools used by the worker. Positional data are extracted from CAD/CAM. Robot path poses are transformed into Cartesian space and validated in simulation, subsequently leading to the generation of robot programs. PathML, an AutomationML-based syntax, integrates robot and manufacturing data across the heterogeneous elements and stages of the manufacturing systems considered. Experiments conducted on the glass adhesive application and welding processes showcased the intuitive nature of the system, with path errors falling within the functional tolerance range.

Figures (10)

• Figure 1: Architecture of the integrated system for intuitive robot programming from human demonstrations and CAD/CAM data. Robot path orientations and velocities are captured through a single-shot human demonstration using magnetic tracking, while positional data are obtained from CAD/CAM. AutomationML-based PathML collects robot data, including poses and velocity, as well as parameters related to glass adhesive application and welding.
• Figure 2: Recorded poses from the magnetic tracker single-shot demonstration following a ground truth rectangular geometry in the x-y plane. The positional error is noticeable, particularly along the z-axis.
• Figure 3: Positional data extracted from CAD/CAM combined with orientation data from the magnetic tracker. The post-processing outputs Euler angles following the x-y-z static convention applied to the robot´s models in use.
• Figure 4: Set of transformations for the glass adhesive application setup (a) and the welding setup (b).
• Figure 5: PathML file containing the welding path data (visualized in the AML Editor software), presented in a CAEX hierarchy tree structure. The hierarchy is organized in a top-down approach, with Layers being the highest level, followed by Tracks and Points. The Points contain the robot pose (position and orientation) data in Cartesian coordinates. The file also stores process parameters.