Adjie Salman

When implementing collaborative robot applications, it is paramount to use a human-centered development approach to ensure a positive user experience and increase acceptance. User Experience (UX) design methods involve validating user experience through evaluations as a basic principle. The earlier UX evaluations are carried out, the greater the added value that can be achieved. For collaborative robot applications, especially those involving large robot systems, these early evaluations are challenging since the entire application will not be available until the final stages of development. The proposed methodological approach to this problem utilizes a rudimentary, scaled test setup for early UX evaluations of the entire robot application, enabling user feedback to be incorporated into the development process early on. The method was applied in a project that developed a collaborative robot application for semi-automated liquid cargo handling in inland navigation. UX assessments were carried out using both the rudimentary test setup and a full-scale prototype in a later development phase. The comparison of both assessments proves the applicability of the proposed methodology. Serving as an overarching framework, this methodology encourages developers to test the entire collaborative robot application in user studies at an early stage, thereby gathering valuable user feedback.

Applications for machine learning (ML) and deep learning (DL) are constantly growing and have already been adopted in the field of particle measurement technology. Even though analytical (ultra-)centrifugation (AC/AUC) is a widely used technique for characterizing dispersed particle systems, ML and DL have not yet been applied in this area. Data evaluation and interpretation in AC/AUC can be challenging and often requires expert knowledge. DL models can help, but their development is limited by a lack of annotated training data. One solution is to generate and use synthetic data instead. In the first part of this study, a model was trained to synthesize data from experiments using a combination of Variational Autoencoder (VAE) and Generative Adversarial Networks (GANs). The results appear highly realistic. Novice users could distinguish real from synthetic samples with only 63% accuracy. Then, a classifier was trained on experimental AC data to categorize real-world examples based on their underlying separation kinetics, testing different DL architectures. After initial training, the models were further fine-tuned with synthetic AC data. ResNet34 models achieved the best performance with 94% accuracy, comparable to an AC expert (91%), while inexperienced users reached only 53%. In the second part of our study, a regression model was trained for the analysis of sedimentation coefficients. Therefore, various generative models were developed and evaluated for synthesizing AUC data based on numerically simulated sedimentation boundaries. The best results were achieved by combining VAE and GAN architectures with embedded physical constraints. However, the generative networks have so far led to additional smearing of the profiles, resulting in a broadening of the sedimentation coefficient distribution and indicating that further refinement is necessary.