Dynamic Robot Trajectory Generation Based on Information from 3D camera

Main functionalities

This module provides a flexible and adaptive way to create robot trajectories dynamically based on point cloud data created automatically with 3D-camera. Module can be utilized in processing of work objects with varying physical characteristics to create robot trajectories dynamically for processes such as painting, sandblasting and pressure washing. Camera may be mounted onto the robot arm or installed in a stationary manner. Secondary functionality is to provide point cloud data of scanned object to be saved as 3D-model file.

Technical specifications

Hardware:
3D camera system (Microsoft Kinect V2, Intel Realsense D435)
A laptop/desktop computer with Windows or Ubuntu 18.04

Software:
AutoMAPPPS
CloudCompare or similar point cloud data processing software

AutoMAPPPS is a family of robot programming software tools developed by Convergent Information Technologies GmbH. The software allows for fast computing of collision-free robot trajectories and program upload to robot controller for execution.
CloudCompare is a 3D point cloud and triangular mesh processing software. The software is originally designed to perform comparison between two dense 3D point clouds or between a point cloud and a triangular mesh. The software has since been developed suitable for more advanced point cloud processing, including algorithms for resampling, scalar fields handling and more.

Inputs and outputs

The input format is depth-map data provided by 3D-camera to CloudCompare and AutoMAPPPS in PLY format. This input data is used to generate a mesh model file of work object in STL or STEP format using CloudCompare, and dynamic robot trajectories are created using AutoMAPPPS.

Formats and standards

Point cloud and 3D model data formats: PLY, STEP, STL

Owner (organization)

Centria University of Applied Sciences https://tki.centria.fi/en

Trainings

Online training material is available through the TRINITY training platform.