This paper attempts to take a comprehensive look at the challenges of representing the spatio-temporal structures and dynamic processes that define a city’s overall characteristics. For the task of urban planning and urban operation, we take the stance that even if the necessary representations of these structures and processes can be achieved, the most important representation of the relevant mindsets of the citizens are, unfortunately, mostly neglected. After a review of major “traditional” urban models of structures behind urban scale, form, and dynamics, we turn to major recent modeling approaches triggered by recent advances in AI that enable multimodal generative models. Some of these models can create representations of geometries, networks and images, and reason flexibly at a human-compatible semantic level. They provide huge amounts of knowledge extracted from huge collections of text and image documents and cover the required rich representation spectrum including geographic knowledge by different knowledge sources, degrees of granularity and scales. We then discuss what these new opportunities mean for coping with the modeling challenges posed by cities, in particular with regard to the role and impact of citizens and their interactions within the city infrastructure. We propose to integrate these possibilities with existing approaches, such as agent-based models, which opens up new modeling spaces including rich citizen models which are able to also represent social interactions. Finally, we put forward some thoughts about a vision of a “social AI in a city ecosystem” that adds relevant citizen models to state-of-the-art structural and process models. This extended city representation will enable urban planners to establish citizen-oriented planning of city infrastructures, to make them into inviting environments that reconcile and foster human culture, city resilience and sustainability.

Extreme heat and air pollution are critical environmental issues that directly impact human health, and their effects are exacerbated under climate change. While numerous studies have investigated the health impacts of extreme heat or air pollution individually, limited research focuses on their combined effects. To address this gap, we constructed an interpretable spatial machine learning model to explore the synergistic interactions between extreme heat and ozone on cancer incidence across 731 urban areas in China. Our model revealed that nighttime extreme heat intensity has stronger association with cancer incidence compared to daytime heat, and that combined exposure to extreme heat and ozone amplifies health risks. We identified significant nonlinear relationships and threshold effects among environmental exposures and health outcomes. Building on these findings, we developed the Urban Dual Environmental Exposure Risk Index (UDEERI) based on IPCC's hazard-exposure-vulnerability framework. UDEERI highlighted significant regional disparities in health risks, identifying high-risk regions primarily in northwestern deserts, central China, eastern coastal areas, and Sichuan-Chongqing metropolitan cluster. Our study fills a critical gap in understanding the compound health risks of extreme heat and elevated ozone under climate change, providing important scientific evidence to support targeted prevention and intervention strategies for sustainable and healthy cities.

An important process in the mining industry is material handling, where trucks are responsible for transporting materials extracted by shovels to different locations within the mine. The decision about the destination of a truck is very important to ensure an efficient material handling operation. Currently, this decision-making process is managed by centralized systems that apply dispatching criteria. However, this approach has the disadvantage of not providing accurate dispatching solutions due to the lack of awareness of potentially changing external conditions and the reliance on a central node. To address this issue, we previously developed a multi-agent system for truck dispatching (MAS-TD), where intelligent agents representing real-world equipment collaborate to generate schedules. Recently, we extended the MAS-TD (now MAS-TDRL) by incorporating learning capabilities and compared its performance with the original MAS-TD, which lacks learning capabilities. This comparison was made using simulated scenarios based on actual data from a Chilean open-pit mine. The results show that the MAS-TDRL generates more efficient schedules.

Digital twins are emerging as a prime analysis, prediction, and control concepts for enabling the Industrie 4.0 vision of cyber-physical production systems (CPPSs). Today’s growing complexity and volatility cannot be handled by monolithic digital twins but require a fundamentally decentralized paradigm of cooperating digital twins. Moreover, societal trends such as worldwide urbanization and growing emphasis on sustainability highlight competing goals that must be reflected not just in cooperating but also competing digital twins, often even interacting in “coopetition”. This paper argues for multi-agent systems (MASs) to address this challenge, using the example of embedding industrial digital twins into an urban planning context. We provide a technical discussion of suitable MAS frameworks and interaction protocols; data architecture options for efficient data supply from heterogeneous sensor streams and sovereignty in data sharing; and strategic analysis for scoping a digital twin systems design among domain experts and decision makers. To illustrate the way still in front of research and practice, the paper reviews some success stories of MASs in Industrie/Logistics 4.0 settings and sketches a comprehensive vision for digital twin-based holistic urban planning.

Though creating socially integrative and sustainable cities is of great interest to many policy makers, urban authorities, public service providers and researchers, how to harness the city population in order to foster such social cohesion, an indispensable part of the process, is a challenge that has yet to be solved. In this chapter, the authors offer a possible solution from the field of natural sciences, viewing cities as living organisms, and demonstrating the use of this principle in a case study of building an online platform, Community of Communities, and how the latter can contribute to the transition towards digital, sustainable, and socially integrative cities in China and Europe. Socially integrative cities are defined as:

Since the Joint Declaration on "The EU-China Partnership on Urbanisation" in 2012, there has been a rapidly growing number of systematic joint research activities on sustainable urbanisation between European and Chinese partners.The "EU-China Sustainable Urbanisation Flagship Initiative" identified four priority areas of mutual interest for EU-China research and innovation collaboration, i.e., sustainable development and urban planning, nature-based solutions for cities, green urban mobility and sustainable energy solutions for cities.Within this framework, the TRANS-URBAN-EU-CHINA research and innovation action started in 2018 with two parallel objectives.On the one hand, it aimed to support policy makers, urban authorities, real estate developers, public service providers and citizens in China to create socially integrative cities in an environmentally friendly and financially viable way.On the other hand, it aimed to help urban stakeholders in Europe to reflect and eventually reconsider their own approaches towards sustainable urbanisation.Real-world methods, instruments and good practice examples from Europe and China, e.g., in terms of social inclusiveness, cultural dynamics, environmental friendliness and economic viability, constituted a basis for comparative analysis.Fourteen project partners of excellence conducted the project.With eight European and six Chinese expert organisations on socially integrative cities, TRANS-URBAN-EU-CHINA combined the best of both worlds to create new insights, practices and role models in sustainable urban development.The Chinese team of partners from government agencies and academia were able to exert a direct impact on society through their national responsibilities for regional and urban planning, research and education.The European partners played a similar role through their positions among European knowledge organisations.The project started from the fact that cities are places of social innovation and engines of economic growth.They attract dynamic groups of society; they provide vast opportunities of interaction, communication and exchange of knowledge; and they thereby lay the foundation for attracting large shares of R&D investment and an innovative service sector.Social integration plays a special role here, as it is directly linked with the economic prosperity of cities, fair access to infrastructure and services, and the fair distribution of wealth and its amenities.This is true for urban development in general, but especially relevant for China as, promoted by various levels of government, the country is transitioning from a less urban to a more urbanised society with increasingly intensified land use and higher quality of life.This book shares the impactful original research results of the project.It is the collaborative product of many stakeholders.It is also among the project's ix main comprehensive academically oriented results.All partners participated in its elaboration in a joint initiative.Mixed author teams, involving European and Chinese experts, are responsible for the individual chapters.Texts were internally reviewed by the editors, as well as further coordinated with the help of the respective work package leaders, who secured additional quality control.In this regard, special thanks go to

Sensors, actuators, machine learning, communication and robotics are paving the way for the introduction of autonomous systems. Autonomous Systems in safety-critical applications require resilient operation of the intended functionality throughout the mission. Especially they must be safe and highly available. However, it is not possible to fully anticipate evolving threats, vulnerabilities and faults during the lifetime of those systems. This requires a resilient systems architecture of the autonomous system. Therefore, a thorough testing and evaluation of such systems is mandatory. In this paper, we present a monkey testing framework for evaluating resilience capabilities of autonomous systems. The framework contains a set of agents with specific role concepts and strategy sets. The framework can be applied to virtual, physical and hybrid testbeds. Due to its modularity the framework is extensible, scalable and also adaptable to different autonomous systems (e.g. mobile robot, manipulator). The monkey testing framework is able to work pseudo-randomized and thus reproducible on a connected system. A logging mechanism annotates the data so that the data can be used for machine learning (e.g. anomaly detection algorithm, selfhealing). We applied the framework on a mobile robotic system in virtual scenarios.

An important logistic process in open-pit mines is material handling due to its high operational costs. In this process shovels extract and load materials that must be transported by trucks to different destinations at the mine. Several centralized systems have been developed to support this process. The methods applied for these systems are based on mathematical programming, heuristic processes or simulation modelling. The main disadvantages in these systems are performing calculations in a timely manner, addressing the dynamics of a mine, and not being able to provide a precise dispatching solution. In this paper, we describe a distributed approach based on Multiagent Systems (MAS). In this approach, the real-world equipment items such as shovels and trucks are represented by intelligent agents. To meet the target in the production plan at minimal cost, the agents must interact with each other. For this interaction, a Contract Net Protocol with a confirmation stage was implemented. To evaluate the MAS, an agent-based simulation with data from a Chilean open-pit mine was used. The results show that the MAS provide more precise solutions than the current centralized systems in a practical calculation timeframe. In addition, the MAS decreases the truck costs by 20% on average.

Material handling is an important process in the mining industry because of its high operational cost. In this process, shovels extract and load materials that must be transported by trucks to different destinations at the mine. When a truck ends an unloading operation, it requires a new loading destination. If a centralized system provides destinations by following dispatching criteria, then one of the main disadvantages of this kind of systems is not being able to provide a precise dispatching solution without knowledge about potentially changed external conditions and the dependency on a central node. In this paper, we describe a distributed approach based on Multiagent Systems (MAS) to alleviate these disadvantages. In this approach, the real-world equipment items such as shovels and trucks are represented by intelligent agents. The agents interact with each other to generate schedules for the machines that they represent. For this interaction, a Contract Net Protocol with a confirmation stage was implemented. In addition, when a machine failure occurs, the agents are able to update their schedules. In order to evaluate the MAS, an agent-based simulation with data from a Chilean open-pit mine was used. The results show that the MAS is able to generate the schedules in a practical computation timeframe. The schedules generated by the MAS decrease the truck cost by 17% on average. Moreover, when a machine failure occurs, the agents are able to repair their schedules in a short period of time.

Software agents in complex, dynamic environments need to update, adapt, and improve their knowledge models for decision making in order to achieve adequate results. Their individual adaption often relies on machine learning from observational data. However, when data is not available in the required quantity and quality, alternative approaches are required. We propose an interaction-based approach to individual model adaption in multiagent systems, describe agent roles and interaction principles and discuss how a goal-oriented transfer of knowledge among agents can he integrated into an agent-based knowledge management framework.

Software agents are a well-established approach for modeling autonomous entities in distributed artificial intelligence. Iterated negotiations allow for coordinating the activities of multiple autonomous agents by means of repeated interactions. However, if several agents interact concurrently, the participants’ activities can mutually influence each other. This leads to poor coordination results. In this paper, we discuss these interrelations and propose a self-organization approach to cope with that problem. To that end, we apply distributed reinforcement learning as a feedback mechanism to the agents’ decision-making process. This enables the agents to use their experiences from previous activities to anticipate the results of potential future actions. They mutually adapt their behaviors to each other which results in the emergence of social order within the multiagent system. We empirically evaluate the dynamics of that process in a multiagent resource allocation scenario. The results show that the agents successfully anticipate the reactions to their activities in that dynamic and partially observable negotiation environment. This enables them to maximize their payoffs and to drastically outperform non-anticipating agents.

The requirements of transport processes have become increasingly complex due to shorter transit times, the individual qualities of shipments, and higher amounts of small sized orders. Especially in courier and express services providing same day deliveries, the high degree of dynamics even increases this complexity. To ensure reliable and flexible planning and control of transport processes, we present a reactive and proactive agent-based system to support the dispatching of logistic transport service providers. Beside the application in simulated real-world processes of our industrial partners, this paper focuses on the impact and relevance of shortest-path queries in the system. We compare the application of state-of-the-art algorithms and investigate the effects of high speed shortest-path computations in agent-based negotiations. The results prove that efficient shortest-path algorithms are an essential key component in agent-based control of dynamic transport processes.

This paper presents an autonomous multiagent system which optimizes the planning and scheduling of industrial processes using the example of courier and express services. In order to handle the rising demands and to capitalize on the increasing optimization potential in transport logistics, which both result from the consequent integration of industrial processes into the Internet of Things and Services, the presented dispAgent solution ensures a flexible, adaptive, and proactive system behavior. Intelligent, selfishly acting agents represent logistic entities, which communicate and negotiate with each other to optimize the allocation of orders to transport facilities. The system has been developed in cooperation with our industrial partner tiramizoo, which is an expert in courier and express services. In order to determine the quality of the computed solutions, we evaluated the system using an established benchmark set and compared the results to best-known solutions. In addition, we further validated the system's performance by multiple simulations of real-world scenarios relying on data which was provided by our industrial partner. The results show that the system achieves high quality solutions for the benchmark set and outperforms a standard dispatching software product in real-world scenarios.

The complexity and dynamics in group age traffic requires flexible, efficient, and adaptive planning and controlling processes. While the general problem refers to the Vehicle Routing Problem (VRP), additional requirements have to be fulfilled in application. Individual properties and priorities of orders, a heterogeneous fleet of vehicles, dynamically incoming orders, unexpected events etc. require a proactive and reactive system behavior. To enable automated dispatching processes, we have implemented a multiagent system where the decision making is shifted from a central system to autonomous, interacting, intelligent agents. To evaluate the approach we used multi agent-based simulation and modeled several scenarios on real world infrastructures with orders provided by our industrial partner. The results reveal that agent-based dispatching meets the increasing requirements in groupage traffic while supporting the combination of pickup and delivery tours and accommodating request priorities, time-windows, as well as capacity constraints.

In this paper we consider efficiently matching logical constraint compositions (called patterns) to noisy observations or to ones which are not well described by existing patterns. The major advantage of our approach to tolerant pattern matching is to exploit existing domain knowledge from an ontological knowledge base represented in description logic in order to handle imprecision in the data and to overcome the problem of an insufficient number of patterns. The matching is defined in a probabilistic framework to ensure that post-processing with probabilistic models is feasible. Additionally, we propose an efficient complete (and optionally approximate) algorithm for this kind of pattern matching. The presented algorithm reduces the number of inference calls to a description logic reasoner. We analyze its worstcase complexity and compare it to a simple algorithm and to a theoretical optimal algorithm.

Digital image processing provides powerful tools for fast and precise analysis of large image data sets in marine and geoscientific applications. Because of the increasing volume of georeferenced image and video data acquired by underwater platforms such as remotely operated vehicles, means of automatic analysis of the acquired image data are required. A new and fast-developing application is the combination of video imagery and mosaicking techniques for seafloor habitat mapping. In this article we introduce an approach to fully automatic detection and quantification of Pogonophora coverage in seafloor video mosaics from mud volcanoes. The automatic recognition is based on textural image features extracted from the raw image data and classification using machine learning techniques. Classification rates of up to 98.86% were achieved on the training data. The approach was extensively validated on a data set of more than 4000 seafloor video mosaics from the Håkon Mosby Mud Volcano.

The learnable evolution model is a stochastic optimization method which employs machine learning to guide the optimization process. LEM3, its newest implementation, combines its machine learning mode with other search operators. The presented research concerns its application within a multi-agent system for autonomous control of container on-carriage operations. Specifically, LEM3 is used by transport management agents that act on behalf of the trucks of a forwarding agency for the planning of individual transport schedules.

Conditional Exponential Models (CEM) are effectively used in several machine learning approaches, e.g., in Conditional Random Fields. Their feature functions are typically either satisfied or not. This paper presents a way to use partially matching feature functions which are satisfied to some degree and corresponding issues while training. Using partially matching feature functions improves the inference accuracy in domains with sparse reference data and avoids overfitting. Unfortunately, the typically used Maximum Likelihood training includes some issues for using partially matching feature functions. In this context three problems (inequality of influence, unlimited weight boundaries and local optima in parameter space) with Improved Iterative Scaling (a popular training algorithm for Conditional Exponential Models) using such feature functions are stated and solved.