Artificial Intelligence meets Multi-Criteria Decision Aiding in smart sustainble cities

在当今的组织中，越来越需要透明度和问责制，特别是涉及环境和公共资源的决定，这对脆弱程度不同和利益冲突的群体有影响。在城市规划和管理方面尤其如此，在城市规划和管理中，饮用水水源保护、废水管理、交通规划、社会住房规划或街道修复等问题必须在人们日益认识到与人类活动有关的社会不平等和环境退化的背景下，在社交媒体、公民参与和大数据时代得到解决。在这种背景下，在2010年代中期，出现了智能可持续城市的概念，即“在信息和通信技术(ICT)的支持下，意识到当地或环境的局限性，在不损害其他人或其他世代满足他们的需求的能力的情况下满足当前居民的需求的城市”(Höjer&Wangle，2015)。但是，一座城市如何才能朝着智能可持续发展的方向发展呢？如何进行公共决策进程，以确保可持续、有力、严格、包容、透明、可接受和公平的决策，同时利用信息和通信技术基础设施？答案是跨学科的，部分在于分析领域，更确切地说，在于参与式多标准决策辅助(MCDA)，这是运筹学的一个分支学科，在过去50年中获得了日益增长的势头，并与人工智能、问题结构方法、地理信息系统和偏好学习等与MCDA决策过程有关的技术相结合。在这个研究计划中，我们将设计一个空间MCDA框架，并开发、验证和评估工具，以支持与智能可持续城市相关的应用程序的MCDA过程：1)什么环境特征、活动、因素和主要困难显著影响MCDA过程的进行，以及如何处理它们？2)如何捕获和考虑所有利益相关者的偏好？3)如何才能获得大量的定量和定性信息，有时称为大数据，被整合到过程中？4)如何评估MCDA过程的质量？5)如何向利益相关者解释和论证决策过程和结果，以确保可接受性？我们的研究项目将按照设计科学方向进行。我们的理论框架将是跨学科的，支撑哲学范式的是批判现实主义，这是一种与我们的多方法论方法一致的哲学。我们将在市政资产管理、饮用水水源地保护和社会住房规划等应用领域应用和测试我们的框架和工具。

In today's organizations, transparency and accountability are increasingly required, especially for decisions involving environmental and public resources, which have an impact on groups with diverse vulnerability levels and conflicting interests. This is particularly true when it comes to city planning and management where issues with drinking water source protection, wastewater management, transportation planning, social housing planning or street rehabilitation must be addressed in a context of increasing awareness of social inequities and environmental degradation related to anthropogenic activities, and in an era of social media, citizens participation and big data. This context has led, in the mid 2010s, to the emergence of the concept of smart sustainable cities, i.e. cities which, "supported by information and communications technologies (ICT) and aware of local or environmental limitations, meet the needs of their current inhabitants without compromising the ability of other people or other generations to meet their needs" (Höjer & Wangel, 2015). But how can a city evolve toward a smart sustainable city? How can public decision processes be conducted to ensure sustainable, robust, rigorous, inclusive, transparent, acceptable and fair decisions while taking advantage of the ICT infrastructure? The answer is interdisciplinary and lies partly in the field of Analytics, more precisely in participatory multicriteria decision aiding (MCDA), a sub-discipline of operations research, which has gained an increasing momentum over the last five decades, and in its combination with techniques from Artificial Intelligence,  Problem Structuring Methods, Geographic Information Systems, and Preference Learning as they pertain to MCDA decision processes. In this research program, we will design a spatial MCDA framework and develop, validate and evaluate tools to support the MCDA process in relation to the following  questions with applications related to smart sustainable cities: 1) What are the environmental characteristics, activities, factors and main difficulties that significantly influence the conduct of a MCDA process and how to deal with them? 2) How can the preferences of all stakeholders be captured and taken into account? 3) How can massive amount of quantitative and qualitative information, sometimes called big data, be integrated in the process? 4) How to evaluate the quality of a MCDA process? 5) How to explain and argue the decision process and the results to the stakeholders to ensure acceptability? Our research program will be conducted following a Design Science orientation. Our theoretical framework will be interdisciplinary and the underpinning philosophical paradigm is critical realism, a philosophy that is consistent with our multimethodological approach. We will apply and test our framework and tools in the following application domains: municipal asset management, drinking water source protection and social housing planning.