Automation in Unstructured Production Environments Using Isaac Sim: A Flexible Framework for Dynamic Robot Adaptability

Abstract

In response to the growing complexity of industrial automation requirements, this paper introduces a comprehensive framework tailored for the automation of industrial robots within unstructured production environments. The framework, emphasizing on adaptability and flexibility, seamlessly merges cutting-edge GPU-based physics engine, the Isaac Sim from Omniverse NVIDIA, with industrial robots, thereby laying the foundation for the development of a robust and versatile digital twin. This digital shadow serves as a main step towards the realization of digital twin technologies in dynamically evolving production environments, facilitating dynamic decision-making processes powered by real-time virtual environmental data. Furthermore, this paper show a compelling application scenario to underscore the practical relevance of the proposed framework. Specifically, the application case centers around a hospital test lab, an onsite facility charged with the preparation of tissue samples for subsequent evaluation by medical professionals. Presently, many of the lab’s tasks are performed manually, underscoring the urgent need for increased automation to enhance efficiency and the working environment. The specific task targeted by this paper involves the re-stacking of microscope slides from a slider fixture to a holder in preparation for subsequent operations. The motivation behind the integration of more dynamic behavior into the robotic system stems from the unstructured nature of incoming samples, coupled with deficiencies in the digital information chain, all within the constraints of a cost-sensitive, non-expert setting. Proving the applicability of this framework in the current test case, it not only enhances efficiency in the hospital test lab scenario but also demonstrates its potential in more advanced applications within the manufacturing field, especially in environments with similar levels of complexity. By removing technical barriers and streamlining the exploration of digital twin applications, this paper contributes to the advancement of automation technologies and sets the stage for future developments in dynamic production environments.