EDAG CityBot – Automated Driving Control of a Multifunctional, Autonomous Robot Vehicle

The EDAG CityBot as part of the “CampusFreeCity” project is a multifunctional, autonomous robot vehicle. The modular ecosystem comprises a driverless tractor vehicle and different utility modules which can be mechanically and electrically coupled to the tractor. These modules are tailored to specific use cases, for instance, to carry cargo in a so-called backpack module or to transport passengers in a trailer module referred to as PeopleMover. In the absence of a physical driver, the CityBot combines a tailored AD sensor platform with a drive-by-wire-system including steer-by-wire and brake-by-wire. The tractor can perform several unique driving maneuvers, such as turn-on-spot, sideways-drive, and diagonal drive. For this purpose, a newly developed chassis system featuring both, single wheel drive and steering was designed. EDAG holistic development approach is further emphasized by the seamless combination of ADAS, chassis and powertrain system to control the CityBot safely in arbitrary driving situations. Following an extensive virtual function development process, all autonomous driving functions and their respective x-by-wire actuators have been modelled and simulated numerically. For this purpose, a new sensor set, consisting of various state-of-the-art technologies, is realized for precise environment detection and mapping. The sensor set is designed in a way that makes safe autonomous driving possible. Sensor fusion and object detection as well as trajectory planning, driving path calculation and the chassis actuation is performed on high performance computers (HPCs) with a self-developed ROS2-based operating system. Furthermore, the CityBot is embedded into a Fleet Management System that sends driving tasks to the vehicle, while also enabling the tele-operation of the tractor, if needed. The major components of the functional architecture of the CityBot are environment perception and ego-localization which enable the creation of a precise map of the vehicle surroundings. A trajectory calculation which follows globally defined routes avoiding obstacles in the driving path and calculating the target steering and velocity values for the acting function. The latter consisting of a state machine model to determine the correct driving mode and, following apart, two separate longitudinal and lateral dynamics controllers as well as a double-track model to determine suitable control parameter values In summary, this contribution outlines the EDAG CityBot’s AD-concept including its requirements, (autonomous) driving functions as well as the underlying sensor and chassis system, the latter two of which are actuated by-wire. The technical introduction of each subsystem also fosters insights of EDAG AD(AS)’ broad development capabilities.