Search

Climate change, global warming, and increasing weatherextremes, especially in Sub-Arctic and Arctic regions with unusual freeze-thaw cycles, cancausemore and more challenges to the infrastructure such as road networks.The maintenanceand repairofroadnetwork can be time consuming andexpensive. Better targeted and proactively planned maintenance could have economical benefits and increase the safeness of theroads. To tackle this, artificial intelligence (AI) and machine learning (ML) techniques with theavailability of digitalised diverse historical and real-time data, can be utilised, on one hand, to betterunderstandthe causes of the thaw damages and frost heave affecting the roads, and on the other hand, to build more advancedforecastingmodels for short- and long-term road conditions and thaw damage risks. In this work, as a first step, for building data-driven ML approaches to Arcticroaddamage forecasting, the possibilitiesof applying different multi-source are analysed.Tothis end, we are applying multi-source data sets of historical weather observations, insitu and mobile measurements of road surface and ground, and response variables of thaw damageand road wearing.As a result, we are showing 1) the benefits of different data sources using explanatory analysis, 2) the importance of different observations explaining the road conditions, and 3) the guideline of building explainable AIandML approaches to combine digitalised information to forecast road conditions such as the thaw damage probability on road network inNorthern Finland., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

The future of smart cities has been significantly impacted by Internet of Things (IoT) and distributed intelligence, where a large scale of data are collected from massive amounts of heterogeneous devices and distributed intelligence brings storage, computing, and Artificial Intelligence (AI) functionality close to the end devices where data are generated for providing novel services and applications. However, AI empowered systems face many challenges due to the inscrutability of complex AI models which weakens the trust of users. This paper provides a general understanding of the underlying concepts and challenges in trustworthy distributed intelligence. A use case of district heating network is illustrated to explore the proposed concepts, technologies, and challenges for enabling trustworthy distributed intelligence for smart cities.