The RoboCup has been an internationally renowned competition for robotics and artificial intelligence for the last 20 years. The aim of the @Work League is to solve robotics problems concerning Industry 4.0 and Smart Factories and to demonstrate the solutions in a lab environment. This is particularly relevant for regional and national industries and is therefore an important component of the FHWS research focus “Digital Production”.


Project Goals

Robotics plays an essential and very diverse role in intelligent manufacturing. In addition to undertaking tasks in production, handling, and assembly, robots are expanding their scope of duties into the area of intralogistics. Components and tools need to be transported between different workstations in a timely manner. To achieve this, so-called driver-less transport systems need to be able to navigate safely in highly dynamic environments; they need to identify objects in their vicinity and pick them up accordingly. In addition, the logistics workflow must be planned efficiently as well as optimized to work hand-in-hand with the manufacturing workflow in order to guarantee a productive manufacturing chain.

SWOT picks an item from the shelf.

The RoboCup@Work League was founded in 2013 to address these diverse research questions through scientific studies. The aim of this international competition is to develop robots that can autonomously perform transportation and handling tasks in an abstract industry setting. The popular RoboCup competition, expanding to five main leagues since starting in 1997, provides an excellent platform for advancing cutting-edge technology in robotics and artificial intelligence in tertiary and secondary education as well as introducing it to the wider public.

Camera-based grasping.

The aim of the project is to establish a RoboCup@Work team at FHWS, thereby addressing several strategic objectives of the university in the areas of research, teaching, and public relations as well as promoting industry collaborations.

SWOT in its natural environment.

The Smart Factory testing arena.

Autonomous navigation.

Omnidirectional movement enabled by mecanum wheels.

Camera-based object recognition and localization.

Point Clouds generated by a depth camera.

Objects extracted from the Point Cloud.

Simulation of the robot and the arena in Gazebo.

The robots’s field of view.

Agile development using Gazebo.

Lucas Reinhart / Team leader

Research fields

Machine Vision

Object Recognition


Neural Networks

Deep Learning

Florian Spieß / Ph.D student

Visiting address

Ignaz-Schön-Straße 11
97421 Schweinfurt



Martin Löser / Laboratory engineer


+49 9721 940-8596


Visiting address

Konrad-Geiger-Straße 2

97421 Schweinfurt



Area(s) of responsibility

Robotics and control technology

Master students

Lukas Kraus

Daniel Blümm

Fabio Mast

Max Dobmann

Prof. Dr. Tobias Kaupp / Supervisor


Visiting address

Ignaz-Schön-Straße 11

97421 Schweinfurt



Office hours

After arrangement

Area(s) of responsibility

Digital Production & Robotics

Picking an aluminium profile.