Robust detection of deformable linear objects (DLOs) is a crucial challenge for the automation handling and assembly cables hoses. The lack training data limiting factor deep-learning-based DLOs. In this context, we propose an automatic image generation pipeline instance segmentation pipeline, user can set boundary conditions to generate industrial applications automatically. A comparison different replication types DLOs shows that modeling as rigid bodies with versatile deformations most...

Industrial part supply often uses universal load carriers. Bin picking offers a variant-flexible option for separating the unordered parts in While numerous approaches exist bin of rigid objects, deformable objects pose challenges. This paper presents concept localization and grasp planning linear (DLOs). First, method reconstruction DLOs is introduced. Then an approach selection uncertainty evaluation described. The proposed promises to identify regions selected DLOs.

Abstract For robust grasping of workpieces with complex surface geometries such as the plugs electrical connectors (ECs), individually designed gripper jaws are commonly required. The manual design latter is time-consuming, iterative, expensive, and requires expert knowledge. Therefore, automating process offers potential to increase efficiency reduce costs. However, ECs often involve interference contours which pose a high risk for grasp fail. Thus, this paper introduces an approach...

During times of flexible production, the human-robot collaboration has potential for automation within assembly process. Optimization manual workstations focuses on workspace layout in particular. The storage boxes are placed grasping area human order to save time. Regarding an application a collaboration, there several criteria finding appropriate task. This paper describes identified regarding which have influence efficient collaboration. These include amongst others movement lengths and...

In industrial production, the handling and mating of electrical connectors are carried out manually in most applications. Therefore, automation offers potential for improvement. Designing a suitable gripper system is crucial step. However, this currently only possible exploratively, requiring empirical knowledge time. This paper presents concept automating design process systems connectors. For purpose, physics simulation mechanical impact cables grasp planning complex geometry considered...

Abstract High product diversity, dynamic market conditions, and a lack of skilled workers are current challenges in manufacturing. Industrial robots autonomously planning completing upcoming production tasks can help companies address these challenges. In this publication, we focus on autonomous task within industrial robotics investigate how to facilitate the use automated techniques from field artificial intelligence for purpose. First, present novel methodology automatically adapt...

Abstract Vision-based robotic picking enables automation of commissioning and sortation disordered parts. To locate parts for grasping, state-of-the-art approaches rely on convolutional neural networks instance segmentation in 2D images. However, this requires sufficiently large training datasets, which are expensive to capture annotate. Therefore, with synthetic data is promising as the can be generated automatically. We present an approach cut-paste method create images industrial use...

Automating the assembly and handling of deformable linear objects requires their robust detection. This paper introduces a new evaluation metric for results from instance segmentation. The enables estimating proportion valid grasp poses graspable specifc gripper models. demonstrate that masks with similar scores in area-based metrics can have different pose validity outcomes. In addition, it is indicated when vacuum gripper, possible to achieve precision recall about 90 %.

The mounting of pre-assembled cables with electrical connectors is mainly carried out manually in industry today. An exemplary application the interconnection battery modules. Automation such assembly tasks offers potential for increasing efficiency but requires design suitable gripper systems. This challenging as cable induces state-dependent forces and torques on system, which must be transmitted via complex surface geometries plugs. Currently, required grasp force cannot determined...