Functional Safety Control

Guarantees by Design

Mission

As automated systems grow more complex and expand to operate autonomously in open environments, ensuring safe operation solely through testing-based system verification becomes increasingly impractical. The Functional Safety Control research group therefore aims to redefine how the functional safety of automation systems is assured: our goal is to reduce the testing effort required for system verification by formally ensuring safety-related specifications in the design process without imposing overly conservative system behavior.

Approach

We pursue this mission by developing safe-by-design methods that make functional safety a formally verifiable property of the system design, rather than a conclusion drawn from testing. Our work spans three tightly connected layers: formal controller synthesis methods that provide robust safety guarantees, safe motion planning algorithms that maintain these guarantees under real-world uncertainty, and compositional verification approaches that establish system-level functional safety through the formal verification of individual automation components and their interfaces. 

What sets our work apart from established robust control and formal verification approaches is the integration of all three layers within a unified control-theoretic framework, grounded in reachability analysis and differential game theory. This enables scalable, formally verified safety across complete automation architectures, aligned with regulatory certification standards such as ISO 26262 and IEC 61508. 

Working with industry partners in the automotive, robotics, and manufacturing sectors, we develop and validate our methods on application domains where ensuring functional safety without imposing overly conservative system behavior is particularly demanding: autonomous vehicles, mobile robots, industrial robots, and nanopositioning actuators.

Publicly Funded Projects

Beyond our industry collaborations, the group is involved in the following publicly funded research projects:

RAMP
Since 2025, the IRS has been part of the publicly funded RAMP project, in which concepts for the use of mobile robotics in modular biotechnological production facilities are developed to replace complex piping systems and reduce energy consumption, cleaning effort, and product losses. The IRS researches automated aseptic coupling and decoupling of production modules, enabling flexible, contamination-free reconfiguration of the facility. Project partners include KIT-IRS, Boehringer Ingelheim, Siemens, SEW-EURODRIVE, Ruhr University Bochum, and Sartorius Stedim Systems.

RepliCar
From 2023 to June 2026, the IRS participated in the three-year, publicly funded RepliCar project, in which a highly accurate reference sensor system for autonomous driving was developed for efficient investigation and validation of series-production sensor systems through a purpose-built data platform and novel test, simulation, and data-processing methods. The IRS was responsible for minimizing the influence of sensor uncertainties on vehicle motion at the motion planning level. Project partners included ANAVS, AKKODIS, FZI, KIT-IHE, KIT-IRS, Porsche, Offenburg University of Applied Sciences, Freudenberg, GTÜ, HighQSoft, RA Consulting, IPG, IAVF, and Wellenzahl.

Werkstromkinematik
From 2021 to 2022, the IRS participated in the two-year, publicly funded FutureFields project "Werkstromkinematik," in which the concept of highly flexible, robot-based manufacturing processes for reconfigurable production was developed. To increase the stiffness and accuracy of a physically coupled multi-robot manufacturing system, the IRS developed methods for trajectory optimization and model predictive control. Project partners included KIT-WBK (project lead), KIT-IPEK and KIT-IAR, as well as IMP at Karlsruhe University of Applied Sciences.

Staff

  

Christopher Bohn

Head of Research Group

Research Interest:
Adaptive Robust Motion Generation through Differential Games

   

Ben-Micha Piscol

Research Associate

Research Interest:
AI-based vehicle control

Andreas Zürcher

Research Associate

Research Interest:
Control of overactuated nano positioning systems

Manuel Hess

Research Associate

Research Interest:
Trajectory planning with consideration of motion sickness

Lorenz Fehn

Research Associate

Research Interest:
Trajectory planning under uncertainty

Jan Riffel

Research Associate

Research Interest:
Functional Safety in Robotics

   

 

Publications


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