• contact@cap-size.at


List of Publications

A Unified Perception Benchmark for Capacitive Proximity Sensing Towards Safe HRC, ICRA 2021

Radar Based Target Tracking and Classification for Efficient Robot Speed Control in Fenceless Environments, IROS 2021

Virtual Radar: Real-Time Millimeter-Wave Radar Sensor Simulation for Perception-Driven Robotics, IEEE Robotics and Automation Letters 2021

robo-gym– An Open Source Toolkit for Distributed DeepReinforcement Learning on Real and Simulated Robots, IROS 2020

Gesture-based Contactless Control of Mobile Manipulators using Capacitive Sensing, IEEE I²MTC 2020

Radar Sensors in Collaborative Robotics: Fast Simulation and Experimental Validation, ICRA 2020

Special Session on Robotics: Tactile & Proximity Sensing, IEEE I²MTC 2020

Proceedings of the 2nd Workshop on Proximity Perception in Robotics, IROS  2019

Proceedings of the 1st Workshop on Proximity Perception in Robotics, IROS 2018


Industrial Use Cases

Despite growing number of industrial processes being robotized or partially robotized, tasks requiring human-level cognition, perception and dexterity remain hard or impossible to automate. Collaborative robotics addresses these challenges by introducing robot coworkers in human-dominated workspaces. In the course of the CapSize project the requirements from the industry for a collaborative work area in terms of safety, usability, flexibility, etc. were elaborated hand in hand with industrial partners. Based on the identified industrial requirements the CapSize team investigated potential practical applications and searched for a suitable use case that focuses on a real world industrial need that can be adequately addressed through human-robot collaboration.







Use Case INNIO Jenbacher – Rocker Arm Preassembly

The rocker arm is an essential part of the cylinder head of INNIO Jenbacher gas engines (see Figure 1). Current assembly process heavily relies on manual work, which limits cycle time, increases the number of waste parts and puts stress on the worker. Increasing the level of automation would solve these issues as well as reduce costs. In general, the use of robots would introduce an increase in speed, precision and quality. However, the rocker arm preassembly is challenging for robots due to required motion complexity which human worker performs with ease. Currently, standard collaborative robots are not suitable to handle this assembly task due to the low speed in collaborative mode. They are not fast enough to achieve the necessary cycle time when operating in collaboration with human worker.


Novel methods and technologies developed within the scope of CapSize (e.g. sensor technologies, algorithms) should provide solutions to the expected challenges and advance the state-of-the art to the point in which the utilization of Human Robot Collaboration (HRC) for the use case scenario is a possibility. The developments should allow higher robot speed during collaborative operation to meet cycle time requirements of the production line. The use case of INNIO Jenbacher – Rocker Arm Preassembly – fulfills all project requirements for a real world industrial use case example in order to demonstrate sensor application scenarios of the CapSize project.


Figure 1:
Individual/single components of the rocker arm (left). The pre-assembled rocker arm (middle) is a part of the cylinder head of a gas engine (right)

Copyrights: JOANNEUM RESEARCH, INNIO Jenbacher