Cooperative Systems

Cooperative Control Systems

Vision

How will automation interact with humans in the future?

How to create synergies between humans and machines in the context of Industry 4.0?

The research group Cooperative Systems develops a framework for modeling and control of interactions between humans and machines. The individual strengths of human and machine are combined to achieve high performance systems, ready to meet future challenges of automatization. The fields of applications are e.g. Advanced Driver Assistance Systems, Robotics, Medical Technologies and Aerospace Engineering.

Cooperative Control Loop

Cooperative Control Loop

Modeling and Identification

The modeling of cooperative systems forms the basis of automation design for cooperative scenarios. In this context, uncertainties in perception and action need to be considered. Furthermore, semantics enable a strategic description of the interaction. Moreover, the identification of human behavior is essential in automation design.

Control Synthesis

Automation design in a cooperative scenario needs to be capable of a dynamic allocation of authority. Furthermore, it requires the ability to negotiate a common goal with the human. One approach to control cooperative systems is based on game theory and Model Predictive Control (MPC). In order to achieve real-time control, motion primitives are examined.

 

Experiments

Motion Tracking is used in various scenarios to measure human motion and validate identification methods.

An Advanced Driving Simulator with haptic feedback human machine interfaces was developed at the IRS. It is used to validate cooperative control methods in the context of advanced driver assistance systems.

A newly developed Ball-on-Plate experiment will be used to apply cooperative identification and control methods in a highly dynamical scenario.

Publicly Funded Projects

IntelliSpine-Recon - In silico planning, robotic training simulation, and outcome-driven learning for complex oncological spine surgery

Project Leader: Balint Varga
Project Partner: Heidelberg University

Surgeries for rare malignant spinal tumors are highly complex due to structures that are difficult to predict and prone to displacement, which is why nearly half of patients experience complications. The IntelliSpine-Recon consortium aims to reduce this risk using purely digital means, without additional interventions. To this end, KIT is developing a patient-specific, interactive in-silico training environment (“human-in-the-loop”) based on ROS2 and integrated with the dVRK platform. Using digital twin data, game-theoretic shared-control algorithms, and dynamic impedance fields, “haptic guardrails” are created that make critical risk zones physically perceptible to the surgeon and enable cooperative human-robot control.
 

AITONOMY - Artificial Intelligence-driven Traffic Optimization and Intelligent Battery Management for Interaction-Aware autnomous Mobility Systems

Project Leader: Balint Varga
https://aitnm.github.io/

The project aims to develop an integrated framework for autonomous, cooperative, and electric vehicles (ACEVs) that addresses two critical and interdependent challenges: ensuring safe and socially aware navigation in complex urban environments and managing battery health and energy efficiency. Easy data exchange is enabled through a unified layer, promoting cooperation between vehicles.

 

Development of a Novel Cooperative Standing Aid for Patients with Neurodegenerative Diseases

Project Leaders: Balint Varga and Julian Schneider

The goal of the collaborative project between ISKO Koch GmbH and KIT is to develop an innovative standing-up aid with autonomous support for patients with neurodegenerative diseases. For the first time, the electromechanical assistance is individually customized by using sensors to analyze the patient’s moment of inertia via motor-side resistance. According to the hypothesis, this measured parameter serves to categorize the severity of the disease. An algorithm evaluates the sensor signals and converts them into movements via actuators. A continuous feedback loop captures and processes these signals to continuously intervene in the movement process and dynamically adapt the assistance to individual needs.

Staff

Balint Varga

Head of Research Group

Research Interest:
Embodied and Human-Centric Teleoperation

   

Christian Braun

Research Associate

Research Interest:
Shared control of heterogenous robot swarms

Julian Schneider

Research Associate

Research Interest:
Multi level connection for the consistent design of cooperative human-machine systems

Philipp Karg

Research Associate

Research Interest:
Modeling and Identification in Cooperative Human-Machine Scenarios

Sean Kille

Research Associate

Research Interest:
 

Karl Handwerker

Research Associate

Research Interest:
 

Felix Thömmes

Research Associate

Research Interest:

Lucas Günther

Research Associate

Research Interest:

   

Publications


2026
2025
2024
2023
2022
2021
2020
2019
2018
2017
2016
2015
2014
2013