SmartWaterway will make autonomous pallet shuttle barges economically viable for urban logistics by combining low-cost onboard sensors with onshore sensors at critical locations (i.e. bridges, locks & bays). The main technical challenges include sensor fusion algorithms to accurately detect and locate obstacles, the fusion of onboard and onshore sensors at critical locations, accurate UWB-localization suitable for outdoor environments, autonomous navigation based on multi-modal data and reliable communication. At the end of the project a demonstrator will be deployed on the urban waterways in Ghent in cooperation with Stad Gent and De Vlaamse Waterweg.
IDLab Gent (DSLab) will bring in its expertise on dynamic navigation (build up in the smartGLAZ and Conamo project) to develop a novel methodology for autonomous navigation of UGCs that takes into account sensor data from the UGC and fuses it with additional data from sensing and localization infrastructure at complex locations on the waterways. In this way, a dynamic 2D navigational map with an underlying network of weighted connections will be constructed that can queried/updated in (near) real-time and can be used to perform accurate automized path planning. Depending on the estimated complexitiy level at each moment in time, the algorithm/map should also be able to scale to different levels of detail (LoD).