Prof. Dr. Doris Aschenbrenner

Re-integration of highly weathered waste plastics into the circular economy, taking into account microplastic output and greenhouse gas emissions

With an interdisciplinary approach and the active participation of industry and civil society, RewitAl pursues the closing of material cycles and the mobilization of recycling in the plastics industry. The focus is on plastics that are exposed to high levels of weathering, which have tended to be neglected to date but are significant in terms of volume, using the example of sports facilities with plastic surfacing systems. The research activities focus on (i) the environmentally friendly deconstruction of the systems, taking into account the distribution of microplastics in the environment, (ii) the analysis of the optimal eligibility, suitability, aptitude of various recycling processes for the various plastic components of the overall system and (iii) the investigation of possible alternatives based on bioplastics. The accompanying and comparative economic and ecological assessment, rating of materials and processes is highly relevant. Together with a transparent description of the material properties, possible application scenarios and business models are developed. Special attention is also paid to the transfer to economy, business and civil society at different levels and via various items in order to actively support the implementation of the innovation and bioeconomy strategy of the state of Baden-Württemberg.

Implementation of practical examples for "Human in Command"

The AI Cockpit project is developing a new approach to human control of AI-based software systems. The aim of the project is to develop an open-source AI cockpit as a digital control unit. This software is being tested in individual application scenarios and its implementation is being prepared.
This cockpit will be accompanied by 1. academic science and scholarshop (allg.) and empirically tested. It represents our response to the "human oversight" required by the EU's AI Act (Art. 14) and supports productive cooperation between humans and machines by allowing users to oversee, control and influence the system's decisions.

Increasing the degree of utilization of plastic waste through AI-based combination of manual sorting and micro-automation

Plastics such as PET or PE are made from crude oil. Due to their excellent properties, they are resistant to many environmental influences and chemicals. However, this makes it difficult for them to decompose naturally. Despite their short useful life, packaging is the largest area of application for plastics, which generates considerable amounts of waste every day. Recycling will play an even more important role in reducing waste, conserving valuable resources and optimizing their reintegration into the material cycle in the future. The mechanical recycling of plastics is mainly carried out using mechanical processes, procedures. This involves sorting, cleaning and processing plastic waste according to type. Purity of type is particularly important for a high recycling rate. However, the sorting process today is very time-consuming and error-prone. An automated sorting system is often preceded or followed by manual sorting or control sorting.

The aim of the RecycleBot project is to increase the recycling rate of plastic waste by optimizing the sorting process using artificial intelligence. This is accompanied by an increased degree of utilization of plastics from packaging and commercial waste. In addition to large automated sorting systems, smaller collaborative robots (RecycleBot) are used for this purpose, which can take over activities that were previously carried out manually. The basis for this work are so-called deep learning (DL) algorithms. They improve the entire recycling process by making a better distinction between the different types of plastic and not only making sorting easier, but also less error-prone. There is also a focus on new and critical plastics and complex composite packaging.

The RecycleBot can increase the recycling rate of plastic-based packaging and commercial waste from the dual system. It is also expected that bioplastics that are bio-based or biodegradable will be increasingly used in the future. These are already being taken into account in the development of the RecycleBot. The development also focuses on human-robot collaboration and the concept of "human supervisory control": Members of staff (pl. staff) employee (pl. employees) from waste management are involved in the development in order to use AI and human experience synergistically. A special field is the human-machine interface for AI visualization to enable controllable self-organizing systems. This means that members of staff (pl. staff), employees (pl. employees) are always able to review and control the system's decisions. In this way, the project contributes to a sustainable circular economy.

Development and evaluation of a human-centered AI system [HCKIS] for use in continuous and discrete production scenarios, using an IoT-based infrastructure and methods of machine learning and artificial intelligence.

The aim of the scheme, enterprise, goal is to intensify cooperation, cooperative venture, collaboration in the field (of) networked artificial intelligence and to create opportunities for cooperation and links to other disciplines at the institution of higher education. As part of, within the framework of the scheme, a joint technology carrier is to be set up, which will be used in particular to demonstrate methods for real-time-based process analysis and quality assurance using artificial intelligence [AI].

Open and flexible industrial networking with a 5G small cell center: Adaptive production and intra-logistics in SMEs

In contrast to previously discussed applications of 5G in production, this project focuses on the establishment of flexible 5G SmallCells with the possibility of federation. The 5G Campus network in nomadic island operation is operated as an open-interface cell center of a heterogeneous, industrial communication and localization network. This enables a dynamically configurable constellation on the store floor, which reflects the reality of small and medium-sized production companies. SME users usually have "high-mix, low-volume" (HMLV) production requirements and need to automate in order to remain competitive. We therefore offer 5G in a package with automation solutions and integrate i) the latest developments on the component side, ii) flexible business models such as "Robotics as a Service" and iii) a combination of 5G with other radio standards while integrating localization solutions. In 5G++ FlexiCell, an SME-centered solution with a test environment for a "5G-enabled" adaptive production of tomorrow will be realized prototypically and pre-competitively in two fieldlabs: i) intra-logistics and ii) adaptive production with lightweight robots.