The societal mission of mitigating air pollution and greenhouse gas emissions are forcing urban agglomerations worldwide strongly greening their urban transportation systems. The global megatrend of urbanization aggravates those challenges by steadily increasing the demand for urban movements of people and goods. Recent research concludes that the autonomous cars propagated in this context carry the risk of significant rebound effects and therefor make the overall societal benefit appear at least doubtful [Fraedrich et al. 2017; Hörl et al. 2019]. Shared autonomous fleets of electrically powered micro-vehicles, on the other hand, have the potential to reduce emissions through their electric powertrains, to avoid traffic jams by substituting of passenger cars, to achieve a high degree of comfort and flexibility compared to the classic car through automated provision and at the same time strengthen public transport as integrated last mile service. At the same time, micro-vehicles, for example in the form of cargo bikes, can be tailored very variably to a specific usage scenario to exploiting further efficiency gains. The authors propose a use case in which an electrified three-wheeled cargo bike, flexibly called to any location at any time, is provided in an automated manner and can be transferred to manual operation after being handed over to the user. After use, the vehicle is released and returns to the depot or is ready for the next request. The separation into automated provision and manual mobility service simplifies the safety concepts and functional safety of the system and thus, from the authors' point of view, increases the realization potential compared to the privat autonomous vehicles (PAV) or shared autonomous vehicles (SAV). The technical implications of this scenario are very similar to those of the autonomous car, but in some cases address significantly different focus, as the article will show. This paper describes the approaches developed during the prototypical realization of the usage scenario and presents proposals for solutions. For this purpose, first relevant requirements are defined, the existing vehicle and sensor concept are described in detail, and solutions for environment perception, prediction, localization, trajectory planning, and interaction design as well as for the confection of the overall logistics system are presented and evaluated in a simulative or experimental manner.

Automated cargo bikes are intended to complement public transportation in a sharing concept and provide an alternative transportation option for people and goods. In highly automated driving without a seated user, real-time trajectory prediction of other road users is crucial for collision avoidance with other motor vehicles or vulnerable road users (VRU). For this purpose, moving obstacles are detected by environmental sensors and classified and tracked using object detection and tracking algorithms. The current and past position data as well as environmental information are used to predict future positions. In this paper, we present several AI-based trajectory prediction models that are specifically suited for this use case. Our focus is not only on the accuracy of trajectory prediction, but additionally on a robust, real-time and practical application. We consider models that can predict the trajectories with position estimation or distributions for position estimation for each time step in the future. For this aim, we present generative network structures based on Conditional Variational Autoencoder (CVAE) in different variants. After training, the models are integrated into our production system and their computation time is determined on the hardware we use.