Coaction between Automobiles and Mobile Robots - Interoperability for Affordable Last Mile Delivery Solutions

Abstract:
This research paper delves into the challenges faced by autonomous delivery robots when navigating public streets and sidewalks, emphasizing the necessity of interaction with various road participants such as vehicles, motorbikes, cycles, and pedestrians. The study focuses on three main areas: the Operating System, with a specific emphasis on ROS/ROS2 (Robot Operating System), simulation solutions like CARLA and AWSIM for testing and validation, and the analysis of different sensor types and stacks (Lidar, Camera, GNNS). The research explores accurate algorithms and robust sensor fusion solutions to ensure the safe and cost-efficient operation of delivery robots. One notable outcome is the integration of a highly accurate (+/-10 cm) navigation stack based on GNNS. Additionally, the research delves into the design and production aspects of developing delivery robots, aiming for a cost-efficient development approach without necessitating serial production. The ultimate goal is to optimize the interaction between delivery robots and other road participants, ensuring a cost-efficient, safe, and robust implementation of a Last-Mile solution.