Publications
Publications
Understanding URDF: A Dataset and Analysis
IEEE Robotics and Automation Letters, Vol. 9, No. 5, pp. 4479-486, 2024
Created and analysed a dataset of 300+ URDFs from various sources including MATLAB, ROS Industrial, Robotics Toolbox, and Drake. Systematic comparison revealed significant discrepancies between URDF files and between different parsers, highlighting inconsistencies in the de-facto robot modeling standard.
• • •
A Review of Unit Level Digital Twin Applications in the Manufacturing Industry
CIRP Journal of Manufacturing Science and Technology, Vol. 45, pp. 162-189, 2023
Detailed literature review on Digital Twins (DTs) in manufacturing, highlighting the challenges of developing DTs of closed systems.
• • •
Understanding URDF: A Survey Based on User Experience
IEEE 19th International Conference on Automation Science and Engineering (CASE), Auckland, New Zealand, 2023
Conducted a user survey with 500+ participants on the use of URDF, its advantages, perceived challenges, and future prospectives. We find that the majority believe URDF will be used in the future, while many were uncertain whether a new format would take over.
• • •
Towards Modular Digital Twins of Robot Systems
2022 IEEE International Conference on Autonomic Computing and Self-Organizing Systems Companion (ACSOS-C), CA, USA, 2022
Proposes a methodology for developing modular Digital Twins (DTs) for robot systems, and presents their shortcomings and open challenges.
• • •
A Modeling Approach for Composed Digital Twins in Cooperative Systems
2023 IEEE 28th International Conference on Emerging Technologies and Factory Automation (ETFA), Sinaia, Romania, 2023
We propose a modeling approach for Digital Twins (DTs) in cooperative systems that provides composition and reusability. This approach extends from information modeling approaches for DTs to include the representation of behavioral models and semantic relationships through an ontology.
• • •
Towards Easy Robot System Integration: Challenges and Future Directions
2022 IEEE/SICE International Symposium on System Integration (SII)*, Narvik, Norway, 2022
Identification of major challenges in robot system integration, including missing standardisation of interfaces and misleading compatibility specifications. Proposes methods for tackling challenges and reducing time and cost of robot system integration.
• • •
Using GitHub Classroom in Teaching Programming
18th CDIO International Conference*, Reykjavík, Iceland, 2022
We used GitHub Classrooms together with Test-Driven-Development for teaching students how to program, allowing them to receive immediate feedback on their programs from the incorporated tests in the assignments.
• • •
Integration Of The Mape-K Loop In Digital Twins
2022 Annual Modeling and Simulation Conference (ANNSIM), CA, USA, 2022
This paper bridges the concepts of Digital Twins (DTs) and "Monitor-Analyze-Plan-Execute over a shared Knowledge" (MAPE-K) by demonstrating an implementation of a MAPE-K loop for an incubator case study within the context of its DT, and discussing some of the limitations of applying the MAPE-K concept to cyber-physical systems.
• • •
AURT: A Tool for Dynamics Calibration of Robot Manipulators
2022 IEEE/SICE International Symposium on System Integration (SII), Narvik, Norway, 2022
AURT is a tool for calibrating robot manipulator dynamics using a command-line interface. The tools requires the Modified Denavit-Hartenberg parameters of the robot and a data sample of the robots joint positions, such that the optimization algorithm can find the best fit for the models of the rigid-body dynamics and the joint dynamics.
• • •
Detecting Faults During Automatic Screwdriving: A Dataset and Use Case of Anomaly Detection for Automatic Screwdriving
Proceedings of the Changeable, Agile, Reconfigurable and Virtual Production Conference and the World Mass Customization & Personalization Conference, Aalborg, Denmark, 2021
In this paper, we present a use case of using ML models for detecting faults during automated screwdriving operations, and introduce a new dataset containing fully monitored and registered data from a Universal Robot and OnRobot screwdriver during both normal and anomalous operations. We illustrate, with the use of two time-series ML models, how to detect faults in an automated screwdriving application.
• • •
A Universal Mechanism for Implementing Functional Mock-up Units
Proceedings of the 11th International Conference on Simulation and Modeling Methodologies, Technologies and Applications - SIMULTECH, 2021
This paper enables easy development of Functional Mockup Units (FMUs) in various languages, such as Python and C#, using a command-line interface. This also enables development of FMUs in Unity for visualizing the simulated system.
• • •
Portable Runtime Environments for Python-based FMUs: Adding Docker Support to UniFMU
Proceedings of 14th Modelica Conference*, Linköping, Sweden, 2021
This papers enables the development of FMUs in Docker using UniFMU, and showcases it on a robotic manipulator example.
• • •
RoboCIM: Towards a Domain Model for Industrial Robot System Configurators
Proceedings of the 15th International Rule Challenge, 7th Industry Track, and 5th Doctoral Consortium, 2021
This paper demonstrates how non-monotonic logic can be used to create a configurator for modular robotic systems, despite incomplete and in some cases incorrect documentation. It is applied to 20 different robotic parts including manipulators, end-effectors, and the coupling device between them.
• • •
A Co-Simulation Based Approach for Developing Safety-Critical Systems
Proceedings of the 18th International Overture Workshop, 2021
Proposes a process for using co-simulation as evidence in safety cases for safety-critical systems, when other methods may be too risky or expensive. For instance, instead of testing with real humans, we could perform co-simulation experiments as a step before performing any physical system tests.
• • •
Visualisation in a Digital Twin Context
The Engineering of Digital Twins, 2024
This chapter describes the potential of visualisation as a service offered by the digital twin of a cyber-physical system. Much of the value of visualisation technology comes from the ability to present stakeholders with a representation of the state and dynamics of the cyber-physical system, augmented with the results of simulation, analysis and prediction services offered by the digital twin. Forms of visualisation are briefly enumerated and compared.
• • •
AURSAD: Universal Robot Screwdriving Anomaly Detection Dataset
arXiv:2102.01409, 2021
A dataset constructed using a robotic screwdriver, specifically the OnRobot Screwdriver, together with a UR3e robotic manipulator. The dataset contains normal operation and different anomalies that were manually constructed, such as damaged screw thread, missing screw, unwanted component in between the screw and the plate, and damaged plate thread. Machine learning models were trained using anomaly detection in attempt to distinguish between faulty vs non-faulty classes of screwdriving.
• • •
Thesis
Enabling Digitalization in Modular Robotic Systems Integration
Aarhus University, 2023
Research on digitalisation of modular robotic systems integration, addressing key challenges including knowledge silos, vendor lock-in, and inefficient integration processes. Contributions include: developing a configurator for robotic systems inspired by PC configuration, comprehensive analysis of the URDF standard (500+ survey participants, 33k+ dataset views), literature review on Digital Twins in manufacturing, and outline of how these approaches reduce integration time and costs.
• • •
A Co-Simulation Based Approach for Developing Safety-Critical Systems
Aarhus University, 2020
Research on developing safety cases for complex systems operating in dynamic environments. Proposes a process incorporating co-simulation for generating safety evidence, with guidelines on when co-simulation is appropriate. Demonstrated through a case study on an autonomous combine harvester.
• • •